Anthropic shipped Claude Opus 4.8, and Simon Willison highlights the unusually restrained release language: a “modest but tangible improvement.” The model keeps most Opus 4.7 pricing and specs, while evaluations suggest it is more likely to flag uncertainty and less likely to ignore flaws in code it wrote. Developer-relevant changes include mid-conversation system messages and a lower prompt-cache minimum of 1,024 tokens.
阿聯酋技術創新研究所(TII)推出全新開源大語言模型 Falcon 180B,擁有 1800 億參數,並在 3.5 兆 Token 的 RefinedWeb 數據集上進行訓練。該模型在 Hugging Face Open LLM 排行榜上名列前茅,性能超越 LLaMA 2 70B,直逼 Google 的 PaLM-2。然而,其龐大的體積也對硬體提出了極高要求,推論至少需要 640GB 顯存(約 8 張 A100 80GB)。