🔴 High Significance
Developer Tools
🔴 Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement Learning — score 95
Sources: huggingface
Large language models (LLMs) typically receive diverse natural language (NL) feedback through interaction with the environment. However, current reinforcement learning (RL) algorithms rely solely on scalar rewards, leaving the rich information in NL feedback underutilized and leading to inefficient
🔴 OpenClaw-RL: Train Any Agent Simply by Talking — score 85
Sources: huggingface
Every agent interaction generates a next-state signal, namely the user reply, tool output, terminal or GUI state change that follows each action, yet no existing agentic RL system recovers it as a live, online learning source. We present OpenClaw-RL, a framework built on a simple observation: next-s
Infrastructure & Compute
🔴 Flash-KMeans: Fast and Memory-Efficient Exact K-Means — score 75
Sources: huggingface
k-means has historically been positioned primarily as an offline processing primitive, typically used for dataset organization or embedding preprocessing rather than as a first-class component in online systems. In this work, we revisit this classical algorithm under the lens of modern AI system des
🟡 Notable
Developer Tools
🟡 LLM2Vec-Gen: Generative Embeddings from Large Language Models — score 65
Sources: huggingface
LLM-based text embedders typically encode the semantic content of their input. However, embedding tasks require mapping diverse inputs to similar outputs. Typically, this input-output is addressed by training embedding models with paired data using contrastive learning. In this work, we propose a no
🟡 In-Context Reinforcement Learning for Tool Use in Large Language Models — score 55
Sources: huggingface
While large language models (LLMs) exhibit strong reasoning abilities, their performance on complex tasks is often constrained by the limitations of their internal knowledge. A compelling approach to overcome this challenge is to augment these models with external tools -- such as Python interpreter
🟡 MA-EgoQA: Question Answering over Egocentric Videos from Multiple Embodied Agents — score 45
Sources: huggingface
As embodied models become powerful, humans will collaborate with multiple embodied AI agents at their workplace or home in the future. To ensure better communication between human users and the multi-agent system, it is crucial to interpret incoming information from agents in parallel and refer to t
🟢 Incremental
Model Releases
🟢 ID-LoRA: Identity-Driven Audio-Video Personalization with In-Context LoRA — score 25
Sources: huggingface
Existing video personalization methods preserve visual likeness but treat video and audio separately. Without access to the visual scene, audio models cannot synchronize sounds with on-screen actions; and because classical voice-cloning models condition only on a reference recording, a text prompt c
🟢 SVG-EAR: Parameter-Free Linear Compensation for Sparse Video Generation via Error-aware Routing — score 5
Sources: huggingface
Diffusion Transformers (DiTs) have become a leading backbone for video generation, yet their quadratic attention cost remains a major bottleneck. Sparse attention reduces this cost by computing only a subset of attention blocks. However, prior methods often either drop the remaining blocks, which in
Developer Tools
🟢 ReMix: Reinforcement routing for mixtures of LoRAs in LLM finetuning — score 35
Sources: huggingface
Low-rank adapters (LoRAs) are a parameter-efficient finetuning technique that injects trainable low-rank matrices into pretrained models to adapt them to new tasks. Mixture-of-LoRAs models expand neural networks efficiently by routing each layer input to a small subset of specialized LoRAs of the la
🟢 Can Large Language Models Keep Up? Benchmarking Online Adaptation to Continual Knowledge Streams — score 15
Sources: huggingface
LLMs operating in dynamic real-world contexts often encounter knowledge that evolves continuously or emerges incrementally. To remain accurate and effective, models must adapt to newly arriving information on the fly. We introduce Online Adaptation to Continual Knowledge Streams(OAKS) to evaluate th
📄 New Papers
| Title | Category | Score | Link |
|---|---|---|---|
| Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement Learning | developer_tool | 214 | Open |
| OpenClaw-RL: Train Any Agent Simply by Talking | developer_tool | 160 | Open |
| Flash-KMeans: Fast and Memory-Efficient Exact K-Means | infrastructure | 85 | Open |
| LLM2Vec-Gen: Generative Embeddings from Large Language Models | developer_tool | 47 | Open |
| In-Context Reinforcement Learning for Tool Use in Large Language Models | developer_tool | 46 | Open |
| Deactivating Refusal Triggers: Understanding and Mitigating Overrefusal in Safety Alignment | cs.AI | 0 | Open |
| Agentic AI for Embodied-enhanced Beam Prediction in Low-Altitude Economy Networks | cs.AI | 0 | Open |
| Stop Listening to Me! How Multi-turn Conversations Can Degrade LLM Diagnostic Reasoning | cs.AI | 0 | Open |
| ARROW: Augmented Replay for RObust World models | cs.AI | 0 | Open |
| Efficient Cross-View Localization in 6G Space-Air-Ground Integrated Network | cs.AI | 0 | Open |
| Entropy Guided Diversification and Preference Elicitation in Agentic Recommendation Systems | cs.AI | 0 | Open |
| Deployment-Time Reliability of Learned Robot Policies | cs.AI | 0 | Open |
| Speak or Stay Silent: Context-Aware Turn-Taking in Multi-Party Dialogue | cs.AI | 0 | Open |
| Contextual Graph Representations for Task-Driven 3D Perception and Planning | cs.AI | 0 | Open |
| Evaluation format, not model capability, drives triage failure in the assessment of consumer health AI | cs.AI | 0 | Open |