Fable 5.2 发布：芯片设计与生物学能力提升，RSI 任务被有意削弱
GDPx 是少数由领域专家出题的基准，能看清 Fable 5.2 到底强在哪。有意思的是 Anthropic 主动压低了 RSI 任务的表现，分数上直接看得出来。如果你拿模型做芯片、生物或写代码相关工作，这份分项数据很有参考价值。
Fable 5.2 在 GDPx 基准测试中的芯片设计、软件工程、生物学和材料科学类别得分显著提升。GDPx 是唯一在这些领域做到细粒度评测的基准，任务由真正的领域专家设计，区别于其他常见基准。
不过该模型并非全面变强：在金融、法律、量化交易、RSI（递归自我改进）和硬件设计等类别上没有提升。文中指出，Anthropic 明确在 RSI 任务上对 Fable 5.2 做了削弱处理，而这一调整直接反映在了 GDPx 的分数上。
这表明头部实验室开始按领域差异化地分配模型能力，在能力扩张与安全边界之间做显式取舍。对用 Fable 5.2 做科研或工程类工作的开发者来说，GDPx 这类专家级基准比通用榜单更能反映实际可用性。
Fable 5.2 is significantly better at chip design, software engineering, biology, and material science.
GDPx is the only eval that test these categories at this level of detail. True experts make these tasks, unlike other benchmarks.
Notably it is not better at finance, legal, quantative trading, RSI, or hardware design.
Anthropic explicitly nerfed this model in RSI tasks and it's interesting seeing that reflected in the score.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Fable 5.2, GDPx, Anthropic, AI安全.
Source attribution: curated from Dylan Patel (SemiAnalysis) (@dylan522p). Cross-source corroboration: 11 independent sources have covered this development (including elvis, Claude, rohanpaul_ai), confirming industry-wide signal strength.
