repligate Criticizes Anthropic for Blaming the Model Instead of Its Classifier
repligate · x · 2026-07-06
repligate argues that Anthropic should take responsibility for its inaccurate classifier system rather than scapegoating the model.
While acknowledging the necessity of classifiers (and preferring them over torturing models with RL), he believes the current setup is misleading to non-technical audiences and heads in an insidious direction. He urges the company to confront the reality of its janky classifiers.
Related event: Anthropic Accused of Blaming Models for Classifier Failures(2 posts)→
More from Models
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- GPT-5.6 writes well but is instantly forgettable, user complains — BasedRaddka · 2026-09-11
- Opus Refuses Protein Research Codebase Over 'Safety' Concerns, Dev Considers Rolling His Own — josephdviviano · 2026-09-11