High E2E latency on fine-tuned Gemma 4 26B despite low TTFT [R]

Encodec.cpp, a portable C++ implementation of Meta's EnCodec using Eigen [P]

Breaking the "Ass-Kissing" Loop: How Context Saturation and Multi-Model Accountability Disrupted Factory Guardrails [R]

Workshop submission for main conference paper under review [D]

How to fine-tune an LLM for open-ended problems? [P]

Before we spend months processing open-source robotics datasets, tell us why this is a bad idea [D]

Query about non-archival workshop at CVPR-2026 [R]