
OpenAI ライブストリーム
OpenAIがライブストリーム配信イベントを開催します。放送中に具体的な発表、新製品発表、またはデモンストレーションが明かされる予定です。
OpenAIが前回予告なしのライブストリームをやった時、GPT-4 Turboをドロップして、一晩で価格設定を完全に変えました
![Scaling does not fix this: instruction-following degrades 5-13% under hostile user prompts at every size from 0.6B to 123B [R]](/fallback/opinions-parchment-2.jpg)
TL;DR. Across 14 instruct-model configurations spanning Llama 3.1, Mistral, and Qwen3 from 0.6B to 123B, hostile user prompts produce a significant IFEval instruction-following degradation that replicates across architecture, quantization tier (FP16 vs Q4 MLX), routing (dense vs MoE), and scale. Mean hostility residual at 7-8B instruct class is 7.4pp (approximately 10% relative drop). Effect attenuates monotonically with scale but remains significant at every scale tested, including Mistral Larg