Latest stories
Manus to Resume Operations as Independent Company, Users May Need to Act
Manus announced it will soon resume operations as an independent company, and some users may need to take action regarding their a...
OpenAI's Daybreak Models Now on AWS Bedrock for Enterprise Security
OpenAI and AWS have partnered to make Daybreak cybersecurity models available through Amazon Bedrock, aiming to bolster enterprise...
OpenAI Begins Testing Ads in ChatGPT, Promising Clear Labels and Privacy
OpenAI has started testing ads in ChatGPT to sustain free access, emphasizing transparency, independence of answers, and user priv...
RoMeRL: A New Method to Balance Feedback and Avoid Memory-Reward Traps in Self-Evolving Agents
Researchers introduce Reduced-Order Memory Reinforcement Learning (RoMeRL), a method that compresses trajectory-indexed memory uti...
RynnValue: Temporal Distance as a Scalable Reward Signal for Robot Learning
Hugging Face highlights RynnValue, an open-source value foundation model that uses temporal distance instead of preferences to lea...
Hugging Face Researchers Unveil Evidence-RL: A New Method to Make VLMs Reason from Visual Evidence
A new paper from Hugging Face introduces Counterfactual Evidence Disentanglement (CED), a training-time audit that ensures vision-...
Evo-Bench: New Benchmark Tests Whether LLMs Can Improve Their Own Agent Harness
Hugging Face researchers introduce Evo-Bench, the first benchmark designed to isolate and evaluate language models' ability to aut...
Business Arena: New Benchmark Reveals LLM Agents Struggle with Realistic Business Operations
A new benchmark from Hugging Face and Accio evaluates LLM agents in a realistic cross-border shop, finding a ninefold gap in perfo...
Interpretability Scales with Capability in New Training-Time Approach
A new paper from Hugging Face shows that making interpretability a training constraint yields scalable, disentangled representatio...
OasisKV: Boosting LLM Throughput by Prefetching Sparse KV Caches Beyond HBM
OasisKV, a new memory-centric inference system from Hugging Face researchers, stores full KV caches in cheaper memory tiers and us...
Hugging Face Research: Three-Stage Framework Boosts Follow-Up Edit Suggestions in Image Conversations
A new multimodal framework from Hugging Face improves follow-up edit suggestions in image-creation conversations, reducing visual...
Researchers Expose Flaw Allowing Theft of Hidden Reasoning from Major AI APIs
A new study reveals that encrypted reasoning traces from proprietary LLMs can be intercepted and decrypted by injecting them into...