Podcast

Daily Paper Cast

Author: Jingwen Liang, Gengyu Wang
Feed episode count: 2000
Official site: https://dailypapercast.transistor.fm/
RSS: https://feeds.transistor.fm/daily-paper-cast-ai
JSON: /v1/public/podcasts/daily-paper-cast-7079649
Markdown: /podcast/daily-paper-cast-7079649.md

Summary

We update every weekday to discuss highest-voted papers from Huggingface Daily Paper (https://huggingface.co/papers). Both the podcast scripts and audio are generated by AI. Feedback and suggestions are welcome! Email us: [email protected] Creator: Jingwen Liang, 3D ML, https://www.linkedin.com/in/jingwen-liang/ Gengyu Wang, LLM ML, http://wanggengyu.com Listen on: Spotify: https://open.spotify.com/show/21nrhmdaA8qoBiH8q03NXL Apple Podcast: https://podcasts.apple.com/us/podcast/daily-paper-cast/id1777620236 Cover Image by Kawen Kuang https://kawen.art

Episodes

Sep 9, 2026 · not_requested

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

Duration seconds: 1197
Source: https://share.transistor.fm/s/abdae4bf

HTML Markdown JSON

Sep 9, 2026 · not_requested

Eliciting Weak-to-Strong Generalization with On-Policy Reverse Distillation

Duration seconds: 1281
Source: https://share.transistor.fm/s/7cb4c5a7

HTML Markdown JSON

Sep 9, 2026 · not_requested

Omni Interaction Agent Technical Report

Duration seconds: 1343
Source: https://share.transistor.fm/s/f38447c2

HTML Markdown JSON

Sep 9, 2026 · not_requested

DriveZero: End-to-End Driving Beyond Human Demonstrations

Duration seconds: 1333
Source: https://share.transistor.fm/s/b7ae44fe

HTML Markdown JSON

Sep 9, 2026 · not_requested

GE-Act 2.0: Pretraining and Scaling a World-Action Model for Robotic Manipulation

Duration seconds: 1206
Source: https://share.transistor.fm/s/38c0118b

HTML Markdown JSON

Sep 9, 2026 · not_requested

Mask Forcing: Improving Autoregressive Video Diffusion Distillation via Dual-Noise Masking Rollout

Duration seconds: 1255
Source: https://share.transistor.fm/s/d98e3d61

HTML Markdown JSON

Sep 9, 2026 · not_requested

AuK Technical Report: An Open-Source Foundational Model for Speech Generation and Editing

Duration seconds: 1171
Source: https://share.transistor.fm/s/9fd46524

HTML Markdown JSON

Sep 9, 2026 · not_requested

Miles v0.1: Production-Level Post-Training

Duration seconds: 1234
Source: https://share.transistor.fm/s/354e197c

HTML Markdown JSON

Sep 9, 2026 · not_requested

SceneMosaic: Efficient and Diverse Simulation-Ready Scene Generation via Hybrid Agentic Layout Evolution

Duration seconds: 1100
Source: https://share.transistor.fm/s/92fa5cfc

HTML Markdown JSON

Sep 4, 2026 · not_requested

LLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes

Duration seconds: 1369
Source: https://share.transistor.fm/s/57ee1150

HTML Markdown JSON

Sep 4, 2026 · not_requested

LatentPress: Context Compression Beyond Text and Vision

Duration seconds: 1222
Source: https://share.transistor.fm/s/d0235eed

HTML Markdown JSON

Sep 4, 2026 · not_requested

Why Gated DeltaNet Survives 4-Bit Quantization: NVFP4 W4A4 for the Recurrent Half of a Hybrid 27B LLM

Duration seconds: 1203
Source: https://share.transistor.fm/s/43af8c54

HTML Markdown JSON

Sep 4, 2026 · not_requested

Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning

Duration seconds: 1237
Source: https://share.transistor.fm/s/b0997aa9

HTML Markdown JSON

Sep 4, 2026 · not_requested

Rethinking On-Policy Distillation of Large Language Models II: One Training Example

Duration seconds: 1332
Source: https://share.transistor.fm/s/e95ac04d

HTML Markdown JSON

Sep 4, 2026 · not_requested

Puffin-World: Scaling a Unified Multimodal Model with Native 3D World States

Duration seconds: 1323
Source: https://share.transistor.fm/s/fa0f3c67

HTML Markdown JSON

Sep 4, 2026 · not_requested

Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments

Duration seconds: 1322
Source: https://share.transistor.fm/s/041c078c

HTML Markdown JSON

Sep 3, 2026 · not_requested

Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills

Duration seconds: 1276
Source: https://share.transistor.fm/s/326e5728

HTML Markdown JSON

Sep 3, 2026 · not_requested

SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models

Duration seconds: 1241
Source: https://share.transistor.fm/s/5023ec64

HTML Markdown JSON

Sep 3, 2026 · not_requested

EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction

Duration seconds: 1265
Source: https://share.transistor.fm/s/b2cc9e6f

HTML Markdown JSON

Sep 3, 2026 · not_requested

It Takes Two to Match: Co-Evolving Generative Retriever with Reinforcement Learning

Duration seconds: 1241
Source: https://share.transistor.fm/s/48c332ae

HTML Markdown JSON

Sep 3, 2026 · not_requested

Language Models Can Control Their Own Attention

Duration seconds: 1259
Source: https://share.transistor.fm/s/c6f3d073

HTML Markdown JSON

Sep 3, 2026 · not_requested

On the Design Fundamentals of Pixel Text Representation Learning

Duration seconds: 1182
Source: https://share.transistor.fm/s/df027c33

HTML Markdown JSON

Sep 2, 2026 · not_requested

StudentSim: Training LLM-based Student Simulators

Duration seconds: 1248
Source: https://share.transistor.fm/s/819545df

HTML Markdown JSON

Sep 2, 2026 · not_requested

SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers

Duration seconds: 1268
Source: https://share.transistor.fm/s/d8dde809

HTML Markdown JSON

Sep 2, 2026 · not_requested

UI-Venus-2 Technical Report

Duration seconds: 1384
Source: https://share.transistor.fm/s/2eb70f45

HTML Markdown JSON

Aug 1, 2026 · not_requested

AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

Duration seconds: 1178
Source: https://share.transistor.fm/s/9a953c88

HTML Markdown JSON

Aug 1, 2026 · not_requested

Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

Duration seconds: 1392
Source: https://share.transistor.fm/s/05c65ebb

HTML Markdown JSON

Aug 1, 2026 · not_requested

Metis: Memory Foundation Model

Duration seconds: 1113
Source: https://share.transistor.fm/s/22649fde

HTML Markdown JSON

Aug 1, 2026 · not_requested

Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering

Duration seconds: 1298
Source: https://share.transistor.fm/s/dd0c00bc

HTML Markdown JSON

Aug 1, 2026 · not_requested

PhiZero: A World Model Built Around Physical Language

Duration seconds: 1177
Source: https://share.transistor.fm/s/9e6fc395

HTML Markdown JSON

Aug 1, 2026 · not_requested

VideoCoCo: Code-as-CoT for Physically-Consistent Video Generation via an Agentic Dual-Engine System

Duration seconds: 1341
Source: https://share.transistor.fm/s/d0fc4240

HTML Markdown JSON

Aug 1, 2026 · not_requested

Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory

Duration seconds: 1343
Source: https://share.transistor.fm/s/758c95f9

HTML Markdown JSON

Aug 1, 2026 · not_requested

Beacon: Knowing When and How to Perform Agentic Visual Reasoning

Duration seconds: 1218
Source: https://share.transistor.fm/s/11578e60

HTML Markdown JSON

Aug 1, 2026 · not_requested

BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms

Duration seconds: 1217
Source: https://share.transistor.fm/s/5a0e9032

HTML Markdown JSON

Aug 1, 2026 · not_requested

Flux-OPD: On-Policy Distillation with Evolving Contexts

Duration seconds: 1238
Source: https://share.transistor.fm/s/c0b436c9

HTML Markdown JSON

Jul 31, 2026 · not_requested

TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM

Duration seconds: 1250
Source: https://share.transistor.fm/s/bae73e74

HTML Markdown JSON

Jul 31, 2026 · not_requested

CoRT: Counterfactual Replay for Token-Level Rubric-Guided Policy Optimization

Duration seconds: 1206
Source: https://share.transistor.fm/s/7b7b9011

HTML Markdown JSON

Jul 31, 2026 · not_requested

HumanCLAW: Can Vision-Language Models Act Through a Body?

Duration seconds: 1181
Source: https://share.transistor.fm/s/6fec1841

HTML Markdown JSON

Jul 31, 2026 · not_requested

DecoEvo: Score-Decoupled Co-Evolution of Solver and Rubric-Generator Skills in Text Space

Duration seconds: 1179
Source: https://share.transistor.fm/s/7a27ea4d

HTML Markdown JSON

Jul 31, 2026 · not_requested

CLBench-V: Evaluating Multimodal Context Learning from Grounding to Knowledge Acquisition

Duration seconds: 1279
Source: https://share.transistor.fm/s/873348af

HTML Markdown JSON

Jul 31, 2026 · not_requested

CAST: Game Solvers as Turn-Level Teachers for LLM Agents

Duration seconds: 1231
Source: https://share.transistor.fm/s/da38c7a9

HTML Markdown JSON

Jul 30, 2026 · not_requested

HiFi-UMI: Learning Deployable Manipulation Policies from High-Fidelity UMI Data Alone

Duration seconds: 1157
Source: https://share.transistor.fm/s/77f7d058

HTML Markdown JSON

Jul 30, 2026 · not_requested

A New Role for Relevance: Guiding Corpus Interaction in Agentic Search

Duration seconds: 1199
Source: https://share.transistor.fm/s/d5b82491

HTML Markdown JSON

Jul 30, 2026 · not_requested

CodeNib: A Multi-View Data System for Serving Repository Context to Coding Agents

Duration seconds: 1140
Source: https://share.transistor.fm/s/35258158

HTML Markdown JSON

Jul 30, 2026 · not_requested

ReDesign: Recovering Editable Design Structures from Images via Agentic Decomposition

Duration seconds: 1291
Source: https://share.transistor.fm/s/2871b6bc

HTML Markdown JSON

Jul 30, 2026 · not_requested

Keep It InMind: Benchmarking the Implicit-Association Blind Spot in Agent Memory

Duration seconds: 1142
Source: https://share.transistor.fm/s/796f31d4

HTML Markdown JSON

Jul 30, 2026 · not_requested

Pass the Baton: Trajectory-Relayed On-Policy Distillation

Duration seconds: 1219
Source: https://share.transistor.fm/s/2c945e76

HTML Markdown JSON

Jul 29, 2026 · not_requested

Kimi K3: Open Frontier Intelligence

Duration seconds: 1234
Source: https://share.transistor.fm/s/0624f2d1

HTML Markdown JSON

Jul 29, 2026 · not_requested

JarvisHub: An Open Harness for Canvas-Native Multimodal Creative Agents

Duration seconds: 1226
Source: https://share.transistor.fm/s/73710d18

HTML Markdown JSON

Jul 29, 2026 · not_requested

Progress Reward Modeling for Robotic Learning: A Comprehensive Survey

Duration seconds: 1298
Source: https://share.transistor.fm/s/ff114da2

HTML Markdown JSON

Jul 29, 2026 · not_requested

From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search

Duration seconds: 1243
Source: https://share.transistor.fm/s/9aa267fb

HTML Markdown JSON

Jul 29, 2026 · not_requested

Rethinking Classifier-Free Guidance in On-Policy Diffusion Distillation

Duration seconds: 1278
Source: https://share.transistor.fm/s/9ffa4904

HTML Markdown JSON

Jul 29, 2026 · not_requested

StateAct: Program State, before Pixels, for Long-Horizon Computer-Use Agents

Duration seconds: 1289
Source: https://share.transistor.fm/s/736ebc34

HTML Markdown JSON

Jul 29, 2026 · not_requested

Data Pyramid for Embodied Manipulation

Duration seconds: 1391
Source: https://share.transistor.fm/s/c02a0306

HTML Markdown JSON

Jul 29, 2026 · not_requested

Sol-Attn: Accelerating Video Generation Inference via On-the-Fly Attention Sparsification

Duration seconds: 1321
Source: https://share.transistor.fm/s/46907ae6

HTML Markdown JSON

Jul 29, 2026 · not_requested

OmniVAE: An Audio-Video VAE with Cross-Modal Alignment for Joint Generation

Duration seconds: 1204
Source: https://share.transistor.fm/s/98c800c7

HTML Markdown JSON

Jul 29, 2026 · not_requested

Oxygen-TryOn: Fashion-Native Foundation Model for Any-item Virtual Try-On

Duration seconds: 1397
Source: https://share.transistor.fm/s/c4cd3577

HTML Markdown JSON

Jul 28, 2026 · not_requested

DataPrep-Bench: Benchmarking LLMs as Training Data Preparators

Duration seconds: 1296
Source: https://share.transistor.fm/s/d4f30016

HTML Markdown JSON

Jul 28, 2026 · not_requested

Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills

Duration seconds: 1166
Source: https://share.transistor.fm/s/9e05d462

HTML Markdown JSON

Jul 28, 2026 · not_requested

Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning

Duration seconds: 1183
Source: https://share.transistor.fm/s/dfaa8e13

HTML Markdown JSON

Jul 25, 2026 · not_requested

AREX: Towards a Recursively Self-Improving Agent for Deep Research

Duration seconds: 1191
Source: https://share.transistor.fm/s/e4ff2020

HTML Markdown JSON

Jul 25, 2026 · not_requested

ReferTrack: Referring Then Tracking for Embodied Visual Tracking

Duration seconds: 1248
Source: https://share.transistor.fm/s/78209b35

HTML Markdown JSON

Jul 25, 2026 · not_requested

K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs

Duration seconds: 1365
Source: https://share.transistor.fm/s/72da1bd9

HTML Markdown JSON

Jul 25, 2026 · not_requested

Visual Contrastive Self-Distillation

Duration seconds: 1284
Source: https://share.transistor.fm/s/36304bfd

HTML Markdown JSON

Jul 25, 2026 · not_requested

Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text

Duration seconds: 1239
Source: https://share.transistor.fm/s/eb4c5b55

HTML Markdown JSON

Jul 24, 2026 · not_requested

SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD

Duration seconds: 1197
Source: https://share.transistor.fm/s/14e3bceb

HTML Markdown JSON

Jul 18, 2026 · not_requested

VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding

Duration seconds: 1481
Source: https://share.transistor.fm/s/bed89c24

HTML Markdown JSON

Jul 18, 2026 · not_requested

SEED: Self-Evolving On-Policy Distillation for Agentic Reinforcement Learning

Duration seconds: 1156
Source: https://share.transistor.fm/s/170dbcfd

HTML Markdown JSON

Jul 18, 2026 · not_requested

LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget

Duration seconds: 1176
Source: https://share.transistor.fm/s/4d98794b

HTML Markdown JSON

Jul 18, 2026 · not_requested

SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration

Duration seconds: 957
Source: https://share.transistor.fm/s/b85efd85

HTML Markdown JSON

Jul 18, 2026 · not_requested

BadWAM: When World-Action Models Dream Right but Act Wrong

Duration seconds: 1263
Source: https://share.transistor.fm/s/abf24a34

HTML Markdown JSON

Jul 18, 2026 · not_requested

KeyFrame-Compass: Towards Comprehensive Evaluation of Keyframe-Conditioned Video Generation

Duration seconds: 1340
Source: https://share.transistor.fm/s/30ad1c1d

HTML Markdown JSON

Jul 18, 2026 · not_requested

MultiRef-Compass: Towards Comprehensive Evaluation of Multi-Reference-to-Audio-Video Generation

Duration seconds: 1251
Source: https://share.transistor.fm/s/019b59c4

HTML Markdown JSON

Jul 18, 2026 · not_requested

Concurrent Image Understanding and Generation: Self-Correcting Coupled Markov Jump Processes

Duration seconds: 1300
Source: https://share.transistor.fm/s/d8d98877

HTML Markdown JSON

Jul 18, 2026 · not_requested

From Pixels to States: Rethinking Interactive World Models as Game Engines

Duration seconds: 1076
Source: https://share.transistor.fm/s/66269037

HTML Markdown JSON

Jul 18, 2026 · not_requested

UniVR: Thinking in Visual Space for Unified Visual Reasoning

Duration seconds: 1097
Source: https://share.transistor.fm/s/d476143c

HTML Markdown JSON

Jul 17, 2026 · not_requested

Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and Editable

Duration seconds: 1152
Source: https://share.transistor.fm/s/cf6127ba

HTML Markdown JSON

Jul 17, 2026 · not_requested

Boogu-Image-0.1: Boosting Open-Source Unified Multimodal Understanding and Generation

Duration seconds: 1168
Source: https://share.transistor.fm/s/5428bda5

HTML Markdown JSON

Jul 17, 2026 · not_requested

Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning

Duration seconds: 1249
Source: https://share.transistor.fm/s/2692e8a7

HTML Markdown JSON

Jul 17, 2026 · not_requested

KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and Skill

Duration seconds: 1336
Source: https://share.transistor.fm/s/da4075d8

HTML Markdown JSON

Jul 17, 2026 · not_requested

OvisOCR2 Technical Report

Duration seconds: 1246
Source: https://share.transistor.fm/s/49b396cf

HTML Markdown JSON

Jul 17, 2026 · not_requested

PolicyShiftGuard: Benchmarking and Improving Policy-Adaptive Image Guardrails

Duration seconds: 1182
Source: https://share.transistor.fm/s/066736ea

HTML Markdown JSON

Jul 17, 2026 · not_requested

MetaView: Monocular Novel View Synthesis with Scale-Aware Implicit Geometry Priors

Duration seconds: 1112
Source: https://share.transistor.fm/s/d0655b61

HTML Markdown JSON

Jul 17, 2026 · not_requested

GigaWorld-Policy-0.5: A Faster and Stronger WAM Empowered by AutoResearch

Duration seconds: 1273
Source: https://share.transistor.fm/s/65d7b1a5

HTML Markdown JSON

Jul 16, 2026 · not_requested

SynthDocBench: Controlled Benchmark for Long-Context Visual Document Understanding

Duration seconds: 1259
Source: https://share.transistor.fm/s/e9f23009

HTML Markdown JSON

Jul 16, 2026 · not_requested

Read It Back: Pretrained MLLMs Are Zero-Shot Reward Models for Text-to-Image Generation

Duration seconds: 1157
Source: https://share.transistor.fm/s/e28356b6

HTML Markdown JSON

Jul 16, 2026 · not_requested

Search Beyond What Can Be Taught: Evolving the Knowledge Boundary in Agentic Visual Generation

Duration seconds: 1164
Source: https://share.transistor.fm/s/d4f85c2a

HTML Markdown JSON

Jul 16, 2026 · not_requested

Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models

Duration seconds: 1179
Source: https://share.transistor.fm/s/2d4e982c

HTML Markdown JSON

Jul 15, 2026 · not_requested

Weak-to-Strong Generalization via Direct On-Policy Distillation

Duration seconds: 1182
Source: https://share.transistor.fm/s/c2959ca2

HTML Markdown JSON

Jul 15, 2026 · not_requested

ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory

Duration seconds: 1338
Source: https://share.transistor.fm/s/dad85862

HTML Markdown JSON

Jul 15, 2026 · not_requested

4D Human-Scene Reconstruction from Low-Overlap Captures

Duration seconds: 1196
Source: https://share.transistor.fm/s/c3142205

HTML Markdown JSON

Jul 9, 2026 · not_requested

RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation

Duration seconds: 1516
Source: https://share.transistor.fm/s/74cad849

HTML Markdown JSON

Jul 9, 2026 · not_requested

RynnWorld-Teleop: An Action-Conditioned World Model for Digital Teleoperation

Duration seconds: 1301
Source: https://share.transistor.fm/s/e56a2721

HTML Markdown JSON

Jul 9, 2026 · not_requested

Hierarchical Sparse Attention Done Right: Toward Infinite Context Modeling

Duration seconds: 1257
Source: https://share.transistor.fm/s/2a4aad18

HTML Markdown JSON

Jul 9, 2026 · not_requested

Vision as Unified Multimodal Generation

Duration seconds: 1578
Source: https://share.transistor.fm/s/b29bc9e3

HTML Markdown JSON

Jul 9, 2026 · not_requested

Gemma 4 Technical Report

Duration seconds: 1431
Source: https://share.transistor.fm/s/d7c796e8

HTML Markdown JSON

Jul 8, 2026 · not_requested

OmniOpt: Taxonomy, Geometry, and Benchmarking of Modern Optimizers

Duration seconds: 1244
Source: https://share.transistor.fm/s/37daef44

HTML Markdown JSON

Jul 8, 2026 · not_requested

UI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning

Duration seconds: 1301
Source: https://share.transistor.fm/s/6d7f3786

HTML Markdown JSON

Jul 8, 2026 · not_requested

ResearchStudio-Reel: Automate the Last Mile of Research from Paper to Poster, Video, and Blog

Duration seconds: 1417
Source: https://share.transistor.fm/s/89017ca7

HTML Markdown JSON

Jul 8, 2026 · not_requested

PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space

Duration seconds: 1421
Source: https://share.transistor.fm/s/3e77b855

HTML Markdown JSON

Jul 8, 2026 · not_requested

ResearchStudio-Idea: An Evidence-Grounded Research-Ideation Skill Suite from ML Conference Outcomes

Duration seconds: 1621
Source: https://share.transistor.fm/s/08b95944

HTML Markdown JSON

Jul 8, 2026 · not_requested

MANCE: Manifold Aware Concept Erasure

Duration seconds: 1340
Source: https://share.transistor.fm/s/b5808a0f

HTML Markdown JSON

Jul 8, 2026 · not_requested

GigaWorld-1: A Roadmap to Build World Models for Robot Policy Evaluation

Duration seconds: 1549
Source: https://share.transistor.fm/s/8e5d08e1

HTML Markdown JSON

Jul 8, 2026 · not_requested

Wan-Streamer v0.2: Higher Resolution, Same Latency

Duration seconds: 1301
Source: https://share.transistor.fm/s/fcd90bf9

HTML Markdown JSON

Jul 8, 2026 · not_requested

EVA-Client: A Unified Data Collection, Inference, and Deployment Framework for Embodied Policies on Real Robots

Duration seconds: 1350
Source: https://share.transistor.fm/s/ad2d40dc

HTML Markdown JSON

Jul 7, 2026 · not_requested

The Mirage of Optimizing Training Policies: Monotonic Inference Policies as the Real Objective for LLM Reinforcement Learning

Duration seconds: 1285
Source: https://share.transistor.fm/s/e36ac227

HTML Markdown JSON

Jul 7, 2026 · not_requested

Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots

Duration seconds: 1350
Source: https://share.transistor.fm/s/0e4399ab

HTML Markdown JSON

Jul 7, 2026 · not_requested

OrbitQuant: Data-Agnostic Quantization for Image and Video Diffusion Transformers

Duration seconds: 1387
Source: https://share.transistor.fm/s/f03d2d1c

HTML Markdown JSON

Jul 7, 2026 · not_requested

VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon

Duration seconds: 1229
Source: https://share.transistor.fm/s/69880cac

HTML Markdown JSON

Jul 7, 2026 · not_requested

DataComp-VLM: Improved Open Datasets for Vision-Language Models

Duration seconds: 1355
Source: https://share.transistor.fm/s/b3d340a7

HTML Markdown JSON

Jul 4, 2026 · not_requested

Program-as-Weights: A Programming Paradigm for Fuzzy Functions

Duration seconds: 1466
Source: https://share.transistor.fm/s/196c54b8

HTML Markdown JSON

Jul 4, 2026 · not_requested

AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents

Duration seconds: 1410
Source: https://share.transistor.fm/s/f25d5edd

HTML Markdown JSON

Jul 4, 2026 · not_requested

EvoPolicyGym: Evaluating Autonomous Policy Evolution in Interactive Environments

Duration seconds: 1386
Source: https://share.transistor.fm/s/f30a665f

HTML Markdown JSON

Jul 4, 2026 · not_requested

Morphing into Hybrid Attention Models

Duration seconds: 1487
Source: https://share.transistor.fm/s/975927dc

HTML Markdown JSON

Jul 4, 2026 · not_requested

AgenticDataBench: A Comprehensive Benchmark for Data Agents

Duration seconds: 1298
Source: https://share.transistor.fm/s/cbf201a6

HTML Markdown JSON

Jul 4, 2026 · not_requested

Multi-Resolution Flow Matching: Training-Free Diffusion Acceleration via Staged Sampling

Duration seconds: 1349
Source: https://share.transistor.fm/s/64025e78

HTML Markdown JSON

Jul 2, 2026 · not_requested

Orca: The World is in Your Mind

Duration seconds: 1473
Source: https://share.transistor.fm/s/328d22cb

HTML Markdown JSON

Jul 2, 2026 · not_requested

Dockerless: Environment-Free Program Verifier for Coding Agents

Duration seconds: 1452
Source: https://share.transistor.fm/s/bbbc53ce

HTML Markdown JSON

Jul 2, 2026 · not_requested

DOPD: Dual On-policy Distillation

Duration seconds: 1526
Source: https://share.transistor.fm/s/ba4a0597

HTML Markdown JSON

Jul 2, 2026 · not_requested

Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models

Duration seconds: 1261
Source: https://share.transistor.fm/s/e917439b

HTML Markdown JSON

Jul 2, 2026 · not_requested

Scenes as Objects, Not Primitives: Instance-Structured 3D Tokenization from Unposed Views

Duration seconds: 1435
Source: https://share.transistor.fm/s/2d8379e9

HTML Markdown JSON

Jul 2, 2026 · not_requested

GEAR: Guided End-to-End AutoRegression for Image Synthesis

Duration seconds: 1590
Source: https://share.transistor.fm/s/a3d4e787

HTML Markdown JSON

Jul 2, 2026 · not_requested

Multi-Block Diffusion Language Models

Duration seconds: 1432
Source: https://share.transistor.fm/s/977ccebb

HTML Markdown JSON

Jul 1, 2026 · not_requested

Agentic Abstention: Do Agents Know When to Stop Instead of Act?

Duration seconds: 1453
Source: https://share.transistor.fm/s/89ee6fbd

HTML Markdown JSON

Jul 1, 2026 · not_requested

LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editing

Duration seconds: 1398
Source: https://share.transistor.fm/s/f9ed41cb

HTML Markdown JSON

Jul 1, 2026 · not_requested

Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent

Duration seconds: 1583
Source: https://share.transistor.fm/s/1744418d

HTML Markdown JSON

Jul 1, 2026 · not_requested

TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents

Duration seconds: 1363
Source: https://share.transistor.fm/s/2777ff4a

HTML Markdown JSON

Jul 1, 2026 · not_requested

Beyond IID: How General Are Tabular Foundation Models, Really?

Duration seconds: 1363
Source: https://share.transistor.fm/s/a625c151

HTML Markdown JSON

Jul 1, 2026 · not_requested

Trimming the Long-Tail of Visual World Modeling Evaluation

Duration seconds: 1554
Source: https://share.transistor.fm/s/91a75afe

HTML Markdown JSON

Jul 1, 2026 · not_requested

AsyncOPD: How Stale Can On-Policy Distillation Be?

Duration seconds: 1358
Source: https://share.transistor.fm/s/f0992ff3

HTML Markdown JSON

Jul 1, 2026 · not_requested

Video-MME-Logical: A Controlled Diagnostic Benchmark for Video Temporal-Logical Reasoning

Duration seconds: 1240
Source: https://share.transistor.fm/s/ff30966b

HTML Markdown JSON

Jul 1, 2026 · not_requested

Bridging VideoQA and Video-Guided Agentic Tasks via Generalized Keyframe Extraction

Duration seconds: 1370
Source: https://share.transistor.fm/s/19acb893

HTML Markdown JSON

Jun 30, 2026 · not_requested

PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation

Duration seconds: 1278
Source: https://share.transistor.fm/s/8100e4d4

HTML Markdown JSON

Jun 30, 2026 · not_requested

Translation as a Bridging Action: Transferring Manipulation Skills from Humans to Robots

Duration seconds: 1436
Source: https://share.transistor.fm/s/db9eff5d

HTML Markdown JSON

Jun 30, 2026 · not_requested

Qwen-Image-2.0-RL Technical Report

Duration seconds: 1630
Source: https://share.transistor.fm/s/e0b171e0

HTML Markdown JSON

Jun 28, 2026 · not_requested

DanceOPD: On-Policy Generative Field Distillation

Duration seconds: 1546
Source: https://share.transistor.fm/s/0a694477

HTML Markdown JSON

Jun 28, 2026 · not_requested

In-Context World Modeling for Robotic Control

Duration seconds: 1403
Source: https://share.transistor.fm/s/6fccae0b

HTML Markdown JSON

Jun 28, 2026 · not_requested

Qwen-Image-Agent: Bridging the Context Gap in Real-World Image Generation

Duration seconds: 1286
Source: https://share.transistor.fm/s/bfdb9a71

HTML Markdown JSON

Jun 28, 2026 · not_requested

OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning

Duration seconds: 1432
Source: https://share.transistor.fm/s/a6ccf440

HTML Markdown JSON

Jun 28, 2026 · not_requested

The Verification Horizon: No Silver Bullet for Coding Agent Rewards

Duration seconds: 1375
Source: https://share.transistor.fm/s/81b0f9d9

HTML Markdown JSON

Jun 28, 2026 · not_requested

ViQ: Text-Aligned Visual Quantized Representations at Any Resolution

Duration seconds: 1552
Source: https://share.transistor.fm/s/8864ed79

HTML Markdown JSON

Jun 13, 2026 · not_requested

EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments

Duration seconds: 1446
Source: https://share.transistor.fm/s/1dcb7985

HTML Markdown JSON

Jun 13, 2026 · not_requested

MiniMax Sparse Attention

Duration seconds: 1562
Source: https://share.transistor.fm/s/d0da285b

HTML Markdown JSON

Jun 13, 2026 · not_requested

SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning

Duration seconds: 1374
Source: https://share.transistor.fm/s/890ea944

HTML Markdown JSON

Jun 13, 2026 · not_requested

InterleaveThinker: Reinforcing Agentic Interleaved Generation

Duration seconds: 1271
Source: https://share.transistor.fm/s/c1b1f49f

HTML Markdown JSON

Jun 13, 2026 · not_requested

FORT-Searcher: Synthesizing Shortcut-Resistant Search Tasks for Training Deep Search Agents

Duration seconds: 1394
Source: https://share.transistor.fm/s/37fe9098

HTML Markdown JSON

Jun 13, 2026 · not_requested

Robust-U1: Can MLLMs Self-Recover Corrupted Visual Content for Robust Understanding?

Duration seconds: 1228
Source: https://share.transistor.fm/s/8c03bbf5

HTML Markdown JSON

Jun 13, 2026 · not_requested

MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling

Duration seconds: 1430
Source: https://share.transistor.fm/s/aa61258c

HTML Markdown JSON

Jun 13, 2026 · not_requested

WeaveBench: A Long-Horizon, Real-World Benchmark for Computer-Use Agents with Hybrid Interfaces

Duration seconds: 1233
Source: https://share.transistor.fm/s/de097270

HTML Markdown JSON

Jun 13, 2026 · not_requested

LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories

Duration seconds: 1395
Source: https://share.transistor.fm/s/0f6333bd

HTML Markdown JSON

Jun 13, 2026 · not_requested

HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers

Duration seconds: 1248
Source: https://share.transistor.fm/s/fd30be8d

HTML Markdown JSON

Jun 11, 2026 · not_requested

ABot-Earth 0.5: Generative 3D Earth Model

Duration seconds: 1367
Source: https://share.transistor.fm/s/5e87e298

HTML Markdown JSON

Jun 11, 2026 · not_requested

Kwai Keye-VL-2.0 Technical Report

Duration seconds: 1532
Source: https://share.transistor.fm/s/3be59bc5

HTML Markdown JSON

Jun 11, 2026 · not_requested

Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution

Duration seconds: 1369
Source: https://share.transistor.fm/s/fdd69514

HTML Markdown JSON

Jun 11, 2026 · not_requested

Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference

Duration seconds: 1269
Source: https://share.transistor.fm/s/c45684dd

HTML Markdown JSON

Jun 11, 2026 · not_requested

SearchSwarm: Towards Delegation Intelligence in Agentic LLMs for Long-Horizon Deep Research

Duration seconds: 1437
Source: https://share.transistor.fm/s/6a886f59

HTML Markdown JSON

Jun 11, 2026 · not_requested

Beyond Uniform Token-Level Trust Region in LLM Reinforcement Learning

Duration seconds: 1589
Source: https://share.transistor.fm/s/dc018303

HTML Markdown JSON

Jun 11, 2026 · not_requested

Flow-DPPO: Divergence Proximal Policy Optimization for Flow Matching Models

Duration seconds: 1300
Source: https://share.transistor.fm/s/3c9dc9d2

HTML Markdown JSON

Jun 11, 2026 · not_requested

SCAIL-2: Unifying Controlled Character Animation with End-to-end In-Context Conditioning

Duration seconds: 1321
Source: https://share.transistor.fm/s/47d67499

HTML Markdown JSON

Jun 11, 2026 · not_requested

Lip Forcing: Few-Step Autoregressive Diffusion for Real-time Lip Synchronization

Duration seconds: 1468
Source: https://share.transistor.fm/s/39e08487

HTML Markdown JSON

Jun 10, 2026 · not_requested

Agents' Last Exam

Duration seconds: 1500
Source: https://share.transistor.fm/s/773034a2

HTML Markdown JSON

Jun 10, 2026 · not_requested

SWE-Explore: Benchmarking How Coding Agents Explore Repositories

Duration seconds: 1387
Source: https://share.transistor.fm/s/75386176

HTML Markdown JSON

Jun 10, 2026 · not_requested

On the Geometry of On-Policy Distillation

Duration seconds: 1613
Source: https://share.transistor.fm/s/1b89b053

HTML Markdown JSON

Jun 10, 2026 · not_requested

LatentSkill: From In-Context Textual Skills to In-Weight Latent Skills for LLM Agents

Duration seconds: 1322
Source: https://share.transistor.fm/s/7f04106c

HTML Markdown JSON

Jun 10, 2026 · not_requested

Latent Spatial Memory for Video World Models

Duration seconds: 1504
Source: https://share.transistor.fm/s/54ecc822

HTML Markdown JSON

Jun 10, 2026 · not_requested

FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention

Duration seconds: 1307
Source: https://share.transistor.fm/s/f0cfc106

HTML Markdown JSON

Jun 10, 2026 · not_requested

CoVEBench: Can Video Editing Models Handle Complex Instructions?

Duration seconds: 1339
Source: https://share.transistor.fm/s/fef810f7

HTML Markdown JSON

Jun 10, 2026 · not_requested

SpatialWorld: Benchmarking Interactive Spatial Reasoning of Multimodal Agents in Real-World Tasks

Duration seconds: 1459
Source: https://share.transistor.fm/s/839d05f7

HTML Markdown JSON

Jun 10, 2026 · not_requested

Human Psychometric Questionnaires Mischaracterize LLM Behavior

Duration seconds: 1508
Source: https://share.transistor.fm/s/30fb1fb6

HTML Markdown JSON

Jun 10, 2026 · not_requested

Echo-Memory: A Controlled Study of Memory in Action World Models

Duration seconds: 1280
Source: https://share.transistor.fm/s/a1d42248

HTML Markdown JSON

Jun 4, 2026 · not_requested

From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain

Duration seconds: 1400
Source: https://share.transistor.fm/s/00f6616a

HTML Markdown JSON

Jun 4, 2026 · not_requested

Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking

Duration seconds: 1540
Source: https://share.transistor.fm/s/7e729012

HTML Markdown JSON

Jun 4, 2026 · not_requested

Trust Region On-Policy Distillation

Duration seconds: 1445
Source: https://share.transistor.fm/s/2bb037b7

HTML Markdown JSON

Jun 4, 2026 · not_requested

KVarN: Variance-Normalized KV-Cache Quantization Mitigates Error Accumulation in Reasoning Tasks

Duration seconds: 1360
Source: https://share.transistor.fm/s/397d6a64

HTML Markdown JSON

Jun 2, 2026 · not_requested

COLLEAGUE.SKILL: Automated AI Skill Generation via Expert Knowledge Distillation

Duration seconds: 1278
Source: https://share.transistor.fm/s/dc215157

HTML Markdown JSON

Jun 2, 2026 · not_requested

Representation Forcing for Bottleneck-Free Unified Multimodal Models

Duration seconds: 1466
Source: https://share.transistor.fm/s/fa685904

HTML Markdown JSON

Jun 2, 2026 · not_requested

Mellum2 Technical Report

Duration seconds: 1302
Source: https://share.transistor.fm/s/562a6bc2

HTML Markdown JSON

Jun 2, 2026 · not_requested

Function2Scene: 3D Indoor Scene Layout from Functional Specifications

Duration seconds: 1318
Source: https://share.transistor.fm/s/7837f3e1

HTML Markdown JSON

Jun 2, 2026 · not_requested

GGT-100K: Generative Ground Truth for Generalizable Real-World Image Restoration

Duration seconds: 1397
Source: https://share.transistor.fm/s/8b7766eb

HTML Markdown JSON

Jun 2, 2026 · not_requested

Towards Streaming Synchronized Spatial Audio Generation via Autoregressive Diffusion Transformer

Duration seconds: 1576
Source: https://share.transistor.fm/s/cea1adc1

HTML Markdown JSON

May 23, 2026 · not_requested

TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation

Duration seconds: 1375
Source: https://share.transistor.fm/s/6639c3a5

HTML Markdown JSON

May 23, 2026 · not_requested

Perception or Prejudice: Can MLLMs Go Beyond First Impressions of Personality?

Duration seconds: 1433
Source: https://share.transistor.fm/s/05a1e45c

HTML Markdown JSON

May 23, 2026 · not_requested

DelTA: Discriminative Token Credit Assignment for Reinforcement Learning from Verifiable Rewards

Duration seconds: 1286
Source: https://share.transistor.fm/s/428a7c58

HTML Markdown JSON

May 23, 2026 · not_requested

$π$-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon Workflows

Duration seconds: 1352
Source: https://share.transistor.fm/s/13d9349a

HTML Markdown JSON

May 23, 2026 · not_requested

Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps

Duration seconds: 1158
Source: https://share.transistor.fm/s/d7961e0a

HTML Markdown JSON

May 23, 2026 · not_requested

ACC: Compiling Agent Trajectories for Long-Context Training

Duration seconds: 1485
Source: https://share.transistor.fm/s/09dad681

HTML Markdown JSON

May 23, 2026 · not_requested

PhysX-Omni: Unified Simulation-Ready Physical 3D Generation for Rigid, Deformable, and Articulated Objects

Duration seconds: 1395
Source: https://share.transistor.fm/s/56e99b00

HTML Markdown JSON

May 23, 2026 · not_requested

LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning

Duration seconds: 1323
Source: https://share.transistor.fm/s/60bce2d7

HTML Markdown JSON

May 23, 2026 · not_requested

Spreadsheet-RL: Advancing Large Language Model Agents on Realistic Spreadsheet Tasks via Reinforcement Learning

Duration seconds: 1345
Source: https://share.transistor.fm/s/55570710

HTML Markdown JSON

May 23, 2026 · not_requested

WorldKV: Efficient World Memory with World Retrieval and Compression

Duration seconds: 1351
Source: https://share.transistor.fm/s/1ceb95e4

HTML Markdown JSON

May 22, 2026 · not_requested

Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining

Duration seconds: 1171
Source: https://share.transistor.fm/s/8e5e5cca

HTML Markdown JSON

May 22, 2026 · not_requested

Mega-ASR: Towards In-the-wild^2 Speech Recognition via Scaling up Real-world Acoustic Simulation

Duration seconds: 1381
Source: https://share.transistor.fm/s/c0e9ca22

HTML Markdown JSON

May 22, 2026 · not_requested

Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos

Duration seconds: 1501
Source: https://share.transistor.fm/s/a48cc988

HTML Markdown JSON

May 22, 2026 · not_requested

IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools

Duration seconds: 1456
Source: https://share.transistor.fm/s/e66ff5df

HTML Markdown JSON

May 21, 2026 · not_requested

When Vision Speaks for Sound

Duration seconds: 1381
Source: https://share.transistor.fm/s/726dcbfe

HTML Markdown JSON

May 21, 2026 · not_requested

Active Learners as Efficient PRP Rerankers

Duration seconds: 1419
Source: https://share.transistor.fm/s/b44a223d

HTML Markdown JSON

May 21, 2026 · not_requested

Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information

Duration seconds: 1377
Source: https://share.transistor.fm/s/3596cc0f

HTML Markdown JSON

May 21, 2026 · not_requested

AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration

Duration seconds: 1419
Source: https://share.transistor.fm/s/ca104a60

HTML Markdown JSON

May 21, 2026 · not_requested

OpenComputer: Verifiable Software Worlds for Computer-Use Agents

Duration seconds: 1482
Source: https://share.transistor.fm/s/bbca1616

HTML Markdown JSON

May 21, 2026 · not_requested

GoLongRL: Capability-Oriented Long Context Reinforcement Learning with Multitask Alignment

Duration seconds: 1476
Source: https://share.transistor.fm/s/dae8be06

HTML Markdown JSON

May 21, 2026 · not_requested

Process Rewards with Learned Reliability

Duration seconds: 1401
Source: https://share.transistor.fm/s/dfadd2c7

HTML Markdown JSON

May 21, 2026 · not_requested

EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL

Duration seconds: 1642
Source: https://share.transistor.fm/s/9f618c24

HTML Markdown JSON

May 21, 2026 · not_requested

CogOmniControl: Reasoning-Driven Controllable Video Generation via Creative Intent Cognition

Duration seconds: 1402
Source: https://share.transistor.fm/s/b64f1aec

HTML Markdown JSON

May 21, 2026 · not_requested

Harnessing LLM Agents with Skill Programs

Duration seconds: 1320
Source: https://share.transistor.fm/s/102f0911

HTML Markdown JSON

May 20, 2026 · not_requested

Code as Agent Harness

Duration seconds: 1523
Source: https://share.transistor.fm/s/3e5d1003

HTML Markdown JSON

May 20, 2026 · not_requested

SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution

Duration seconds: 1371
Source: https://share.transistor.fm/s/6a5c9833

HTML Markdown JSON

May 20, 2026 · not_requested

LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation

Duration seconds: 1345
Source: https://share.transistor.fm/s/ef0f6ffc

HTML Markdown JSON

May 20, 2026 · not_requested

Lance: Unified Multimodal Modeling by Multi-Task Synergy

Duration seconds: 1391
Source: https://share.transistor.fm/s/f20fd799

HTML Markdown JSON

May 20, 2026 · not_requested

AI for Auto-Research: Roadmap & User Guide

Duration seconds: 1342
Source: https://share.transistor.fm/s/0f48e7af

HTML Markdown JSON

May 20, 2026 · not_requested

CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?

Duration seconds: 1390
Source: https://share.transistor.fm/s/93c04a08

HTML Markdown JSON

May 20, 2026 · not_requested

KVPO: ODE-Native GRPO for Autoregressive Video Alignment via KV Semantic Exploration

Duration seconds: 1420
Source: https://share.transistor.fm/s/14186699

HTML Markdown JSON

May 19, 2026 · not_requested

CiteVQA: Benchmarking Evidence Attribution for Trustworthy Document Intelligence

Duration seconds: 1390
Source: https://share.transistor.fm/s/95291574

HTML Markdown JSON

May 19, 2026 · not_requested

PhysBrain 1.0 Technical Report

Duration seconds: 1524
Source: https://share.transistor.fm/s/053171c2

HTML Markdown JSON

May 19, 2026 · not_requested

MMSkills: Towards Multimodal Skills for General Visual Agents

Duration seconds: 1360
Source: https://share.transistor.fm/s/457d4957

HTML Markdown JSON

May 19, 2026 · not_requested

DexJoCo: A Benchmark and Toolkit for Task-Oriented Dexterous Manipulation on MuJoCo

Duration seconds: 1392
Source: https://share.transistor.fm/s/8743282c

HTML Markdown JSON

May 19, 2026 · not_requested

Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding

Duration seconds: 1272
Source: https://share.transistor.fm/s/e1a410e1

HTML Markdown JSON

May 19, 2026 · not_requested

InsightTok: Improving Text and Face Fidelity in Discrete Tokenization for Autoregressive Image Generation

Duration seconds: 1462
Source: https://share.transistor.fm/s/f4eaa239

HTML Markdown JSON

May 19, 2026 · not_requested

Flash-GRPO: Efficient Alignment for Video Diffusion via One-Step Policy Optimization

Duration seconds: 1264
Source: https://share.transistor.fm/s/48be6e53

HTML Markdown JSON

May 19, 2026 · not_requested

Nudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVR

Duration seconds: 1267
Source: https://share.transistor.fm/s/7f34b019

HTML Markdown JSON

May 16, 2026 · not_requested

Achieving Gold-Medal-Level Olympiad Reasoning via Simple and Unified Scaling

Duration seconds: 1370
Source: https://share.transistor.fm/s/08cf43f7

HTML Markdown JSON

May 16, 2026 · not_requested

Causal Forcing++: Scalable Few-Step Autoregressive Diffusion Distillation for Real-Time Interactive Video Generation

Duration seconds: 1410
Source: https://share.transistor.fm/s/f653d882

HTML Markdown JSON

May 16, 2026 · not_requested

Self-Distilled Agentic Reinforcement Learning

Duration seconds: 1491
Source: https://share.transistor.fm/s/896103fd

HTML Markdown JSON

May 16, 2026 · not_requested

MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models

Duration seconds: 1617
Source: https://share.transistor.fm/s/25b78099

HTML Markdown JSON

May 16, 2026 · not_requested

SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer

Duration seconds: 1332
Source: https://share.transistor.fm/s/ab72c767

HTML Markdown JSON

May 16, 2026 · not_requested

MemEye: A Visual-Centric Evaluation Framework for Multimodal Agent Memory

Duration seconds: 1370
Source: https://share.transistor.fm/s/88be608c

HTML Markdown JSON

May 16, 2026 · not_requested

Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning

Duration seconds: 1402
Source: https://share.transistor.fm/s/0f4c1b96

HTML Markdown JSON

May 16, 2026 · not_requested

Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems

Duration seconds: 1312
Source: https://share.transistor.fm/s/88a6b0fa

HTML Markdown JSON

May 16, 2026 · not_requested

STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?

Duration seconds: 1411
Source: https://share.transistor.fm/s/ad2fff7b

HTML Markdown JSON

May 16, 2026 · not_requested

WildClawBench: A Benchmark for Real-World, Long-Horizon Agent Evaluation

Duration seconds: 1473
Source: https://share.transistor.fm/s/e8cb6ecf

HTML Markdown JSON

May 15, 2026 · not_requested

MinT: Managed Infrastructure for Training and Serving Millions of LLMs

Duration seconds: 1432
Source: https://share.transistor.fm/s/a9dd0ae7

HTML Markdown JSON

May 15, 2026 · not_requested

MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image

Duration seconds: 1453
Source: https://share.transistor.fm/s/0c7a3b1e

HTML Markdown JSON

May 15, 2026 · not_requested

AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation

Duration seconds: 1410
Source: https://share.transistor.fm/s/ac4e4aa7

HTML Markdown JSON

May 15, 2026 · not_requested

Training Long-Context Vision-Language Models Effectively with Generalization Beyond 128K Context

Duration seconds: 1385
Source: https://share.transistor.fm/s/fab16fc9

HTML Markdown JSON

May 15, 2026 · not_requested

EVA-Bench: A New End-to-end Framework for Evaluating Voice Agents

Duration seconds: 1519
Source: https://share.transistor.fm/s/3a90cf54

HTML Markdown JSON

May 15, 2026 · not_requested

Predicting Decisions of AI Agents from Limited Interaction through Text-Tabular Modeling

Duration seconds: 1499
Source: https://share.transistor.fm/s/3a7cb92e

HTML Markdown JSON

May 15, 2026 · not_requested

Qwen-Image-VAE-2.0 Technical Report

Duration seconds: 1462
Source: https://share.transistor.fm/s/12f43ae7

HTML Markdown JSON

May 15, 2026 · not_requested

TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Tracking

Duration seconds: 1406
Source: https://share.transistor.fm/s/7a4f5ed0

HTML Markdown JSON

May 15, 2026 · not_requested

Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling

Duration seconds: 1418
Source: https://share.transistor.fm/s/46539574

HTML Markdown JSON

May 15, 2026 · not_requested

Many-Shot CoT-ICL: Making In-Context Learning Truly Learn

Duration seconds: 1440
Source: https://share.transistor.fm/s/8b97cdae

HTML Markdown JSON

May 14, 2026 · not_requested

MemPrivacy: Privacy-Preserving Personalized Memory Management for Edge-Cloud Agents

Duration seconds: 1467
Source: https://share.transistor.fm/s/7e697f9d

HTML Markdown JSON

May 14, 2026 · not_requested

SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture

Duration seconds: 1532
Source: https://share.transistor.fm/s/d8453b62

HTML Markdown JSON

May 14, 2026 · not_requested

$δ$-mem: Efficient Online Memory for Large Language Models

Duration seconds: 1471
Source: https://share.transistor.fm/s/979c7e38

HTML Markdown JSON

May 14, 2026 · not_requested

RubricEM: Meta-RL with Rubric-guided Policy Decomposition beyond Verifiable Rewards

Duration seconds: 1357
Source: https://share.transistor.fm/s/99f378e6

HTML Markdown JSON

May 14, 2026 · not_requested

Do Enterprise Systems Need Learned World Models? The Importance of Context to Infer Dynamics

Duration seconds: 1361
Source: https://share.transistor.fm/s/b3c915cd

HTML Markdown JSON

May 14, 2026 · not_requested

World Action Models: The Next Frontier in Embodied AI

Duration seconds: 1483
Source: https://share.transistor.fm/s/679ec15d

HTML Markdown JSON

May 14, 2026 · not_requested

Beyond the Last Layer: Multi-Layer Representation Fusion for Visual Tokenization

Duration seconds: 1520
Source: https://share.transistor.fm/s/f342417e

HTML Markdown JSON

May 14, 2026 · not_requested

Efficient Pre-Training with Token Superposition

Duration seconds: 1475
Source: https://share.transistor.fm/s/4aff0510

HTML Markdown JSON

May 14, 2026 · not_requested

AlphaGRPO: Unlocking Self-Reflective Multimodal Generation in UMMs via Decompositional Verifiable Reward

Duration seconds: 1439
Source: https://share.transistor.fm/s/ee158ab8

HTML Markdown JSON

May 14, 2026 · not_requested

MCP-Cosmos: World Model-Augmented Agents for Complex Task Execution in MCP Environments

Duration seconds: 1304
Source: https://share.transistor.fm/s/7379ec2d

HTML Markdown JSON

May 13, 2026 · not_requested

Qwen-Image-2.0 Technical Report

Duration seconds: 1383
Source: https://share.transistor.fm/s/8130b403

HTML Markdown JSON

May 13, 2026 · not_requested

Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs

Duration seconds: 1439
Source: https://share.transistor.fm/s/1d6bb954

HTML Markdown JSON

May 13, 2026 · not_requested

CollabVR: Collaborative Video Reasoning with Vision-Language and Video Generation Models

Duration seconds: 1526
Source: https://share.transistor.fm/s/855e569e

HTML Markdown JSON

May 13, 2026 · not_requested

TMAS: Scaling Test-Time Compute via Multi-Agent Synergy

Duration seconds: 1397
Source: https://share.transistor.fm/s/3967f7ba

HTML Markdown JSON

May 13, 2026 · not_requested

PaperFit: Vision-in-the-Loop Typesetting Optimization for Scientific Documents

Duration seconds: 1375
Source: https://share.transistor.fm/s/2be242df

HTML Markdown JSON

May 13, 2026 · not_requested

Model Merging Scaling Laws in Large Language Models

Duration seconds: 1304
Source: https://share.transistor.fm/s/dc79b8ed

HTML Markdown JSON

May 13, 2026 · not_requested

SEIF: Self-Evolving Reinforcement Learning for Instruction Following

Duration seconds: 1287
Source: https://share.transistor.fm/s/e7d9944e

HTML Markdown JSON

May 13, 2026 · not_requested

WorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors

Duration seconds: 1346
Source: https://share.transistor.fm/s/60f7d5d5

HTML Markdown JSON

May 13, 2026 · not_requested

Memory-Efficient Looped Transformer: Decoupling Compute from Memory in Looped Language Models

Duration seconds: 1349
Source: https://share.transistor.fm/s/51524a66

HTML Markdown JSON

May 12, 2026 · not_requested

Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers

Duration seconds: 1313
Source: https://share.transistor.fm/s/a62deee0

HTML Markdown JSON

May 12, 2026 · not_requested

Flow-OPD: On-Policy Distillation for Flow Matching Models

Duration seconds: 1575
Source: https://share.transistor.fm/s/56fff24a

HTML Markdown JSON

May 12, 2026 · not_requested

HyperEyes: Dual-Grained Efficiency-Aware Reinforcement Learning for Parallel Multimodal Search Agents

Duration seconds: 1521
Source: https://share.transistor.fm/s/3d74d9c7

HTML Markdown JSON

May 12, 2026 · not_requested

Anisotropic Modality Align

Duration seconds: 1374
Source: https://share.transistor.fm/s/3dadc4f5

HTML Markdown JSON

May 12, 2026 · not_requested

Beyond Retrieval: A Multitask Benchmark and Model for Code Search

Duration seconds: 1274
Source: https://share.transistor.fm/s/7e0e20ae

HTML Markdown JSON

May 9, 2026 · not_requested

MiA-Signature: Approximating Global Activation for Long-Context Understanding

Duration seconds: 710
Source: https://share.transistor.fm/s/4af916c3

HTML Markdown JSON