Veille quotidienne — 22/10/2026
Veille RSS
- A Coding Guide to Cactus Whistle: Does a 16.9 MB Speech Model Hold Up Against Whisper Base? Word Error Rate, Speed, Streaming, Timestamps, Keywords, and Speech-to-Tool Calls on a CPU https://www.marktechpost.com/2026/10/21/cactus-whistle-vs-whisper-base-speech-to-text-benchmark-cpu-guide/
- GaP-MoE : Explicit Expert Specialization in Mixture-of-Experts via Orthogonal Gradient Accumulation and Projection https://arxiv.org/abs/2610.29355
- SimHack: Predicting the onset of detectable reward hacking before RL training https://arxiv.org/abs/2610.28707
- CSD: Common-Support Distillation for Cross-Tokenizer Knowledge Transfer https://arxiv.org/abs/2610.28895
- A Coding Guide to Cactus Whistle: Does a 16.9 MB Speech Model Hold Up Against Whisper Base? Word Error Rate, Speed, Streaming, Timestamps, Keywords, and Speech-to-Tool Calls on a CPU https://www.marktechpost.com/2026/10/21/cactus-whistle-vs-whisper-base-speech-to-text-benchmark-cpu-guide/
- GaP-MoE : Explicit Expert Specialization in Mixture-of-Experts via Orthogonal Gradient Accumulation and Projection https://arxiv.org/abs/2610.29355
- SimHack: Predicting the onset of detectable reward hacking before RL training https://arxiv.org/abs/2610.28707
- CSD: Common-Support Distillation for Cross-Tokenizer Knowledge Transfer https://arxiv.org/abs/2610.28895
Comments