Haiku 5.5 Fails the Lava Lamp Test
Summary
An article examining Anthropic's Haiku 5.5 model and its poor performance on the 'Lava Lamp Test,' an evaluation probing model behavior/reasoning. It highlights shortcomings in a smaller, speed-focused model class.
Similar Articles
Sonnet 5.5 and Haiku 5.5 Coming Soon!
Anthropic is set to release Sonnet 5.5 and Haiku 5.5, which are upcoming updates to their AI language models.
@bcherny: Haiku 5.5 is a good Haiku. With 100k token context, it's also 10x cheaper than Haiku 4.5
Anthropic 发布 Claude Haiku 5.5，宣称是其最便宜、最快且最强的小模型，相比 Haiku 4.5 运行成本降低约 75%，并支持 100k token 上下文窗口。
Claude Haiku 5.5
Anthropic released Claude Haiku 5.5, a fast low-cost model priced at $0.10/$0.50 up to 100K tokens to match GPT-6 Luna, though a less generous tokenizer and a 5x price jump beyond 100K tokens make Luna the better deal for long contexts.
Anthropic tests Fable 5.2 and Opus 5.5 ahead of the release (3 minute read)
Anthropic is quietly testing its upcoming AI models, Fable 5.2 and Opus 5.5, with early user examples showing advanced capabilities in generating interactive JavaScript games and animations without external tools.
Claude Haiku 5.5
Anthropic が Claude Haiku 5.5 をリリースした小型・高速モデルに関するニュース。