The study evaluates Qwen3-0.6B and Qwen3-1.7B to compare weight quantization with option-o…
SteerDistill is a quantization-aware distillation method designed for low-bit LLM agents. …
Apple Machine Learning Research presented a study at a NeurIPS 2026 workshop examining whe…
EVAThink addresses a central efficiency problem in video-language models: under a fixed fr…
MMDM introduces a large-scale multitask benchmark for System-1 decision models, comprising…
Microsoft Research has disclosed Lazarus, a model-weight backdoor that requires no retrain…
Anthropic has unveiled a broader Claude office suite for knowledge workers, including Clau…
Google is rolling out Gemini 4 Carbon to the general public following a two-week restricte…
Adobe is adding AI agents and creative tooling across Photoshop, Premiere, and Firefly, in…
Google Cloud has launched Gemini for Healthcare, a healthcare-focused offering designed to…
Google DeepMind introduced AlphaProof Code, an agentic system for deep formal verification…