HuggingFace
FlexiSLM: A Dynamic and Controllable Frame Rate Spoken Language Model
Spoken language models (SLMs) extend LLMs to speech input and output. Existing SLMs represent speech at fixed frame r...
🤖 AI 前沿资讯
每天自动聚合 AI 领域最新动态
HuggingFace
QVal: Cheaply Evaluating Dense Supervision Signals for Long-Horizon LLM Agents
LLM agents increasingly act over long horizons, where a single trajectory can contain hundreds or thousands of action...
HuggingFace
Unlocking the Visual Record of Materials Science: A Large-Scale Multimodal Dataset from Scientific Literature
The materials science literature encodes decades of experimental knowledge in figures, yet this visual record remains...
HuggingFace
Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing
Existing instruction-based video editing datasets commonly focus on single-task appearance editing, failing to meet t...
HuggingFace
MOPD: Multi-Teacher On-Policy Distillation for Capability Integration in LLM Post-Training
Modern large language models (LLMs) rely on reinforcement learning during post-training to push specific capabilities...
HuggingFace
Lexical Consensus: Grounded Word Learning and Shared Meaning in Artificial Agents
Artificial intelligence systems are commonly evaluated through task performance and behavioral imitation, but such ev...
HuggingFace
Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models
Embodied Vision-Language-Action (VLA) models are typically obtained by fine-tuning powerful pretrained VLMs on roboti...
HuggingFace
Are We Measuring Strategy or Phrasing? The Gap Between Surface- and Approach-Level Diversity in LLM Math Reasoning
Diversity in LLM mathematical reasoning is critical for exploration, but common diversity metrics mostly capture surf...
HuggingFace
Hierarchical Experimentalist Agents
Large language models (LLMs) are increasingly used to take actions in the real world and support human decision-makin...
HuggingFace
Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?
Multi-fingered robots promise the speed and dexterity of human hands, yet challenging problems such as precise assemb...
HuggingFace
SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions
We introduce SWE-Interact, a new testbed for evaluating coding agents on multi-turn, interactive, user-driven softwar...
HuggingFace
TRIAGE: Role-Typed Credit Assignment for Agentic Reinforcement Learning
Agentic reinforcement learning requires assigning credit to environment-facing actions such as searches, clicks, edit...