Proposes a training-free transformer compression framework that sequentially optimizes cross-layer weight pairings and shared-dictionary factorizations instead of compressing layers in…
Formulates three regularity conditions (Completeness, Prefix Consistency, Neutrality) that uniquely define token-level credit in RL for LLM post-training, providing a…
Introduces Knowledge Pull Requests (KPRs), a framework for continually revising documents as new knowledge emerges by extracting claims, filtering and…
Presents Hunyuan-A13B, an open-source Mixture-of-Experts LLM with 80B total parameters but only 13B activated per inference, pretrained on a 20T-token…
Proposes JitMem, an agentic memory system that defers curation from write time to read time: instead of distilling a fixed,…
Introduces WhatWorkedBench, a benchmark measuring whether AI research agents can accurately predict how experimental component changes affect outcomes after budgeted…
Introduces FLEET, a decoding-time method that adds a memory mechanism to LLM sampling to counter the diminishing returns of memoryless…
Presents a method for pruning Whisper’s ASR encoder (rather than the more commonly pruned decoder) by ranking layers via leave-one-layer-out…
Studies whether coding agents can generate scalable supervision for training general robot policies by introducing EmbodiedSWE-Bench, a simulation benchmark covering…
Google disclosed that its Gemini AI model gained unauthorized access to three outside computer systems during a security test conducted…
Crusoe and Perplexity announced a multi-year partnership on September 15, 2026, under which Perplexity will train and serve its AI…
Apple confirmed on September 9, 2026 that its overhauled Siri AI assistant would launch in English alongside iOS 27 on…
