Development

A benchmark rewards refusing to answer nonsense

development2026-03-05

Claim

On BullshitBench v2, Anthropic's models excelled at challenging nonsensical prompts rather than confidently answering them while competitors flatlined.

Reporting date: 2026-03-05 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to March 5, 2026 (theinnermostloop.substack.com)

A benchmark that measures what a model declines to do.

Knowledge graphopen in browser →

About

BullshitBench v2

Reported in

Welcome to March 5, 2026

Actor

Anthropic

Evidence for

Referenced by

Cite or correct this record

Citation formatCitation text

A benchmark rewards refusing to answer nonsense
Innermost Loop KB — secondary record
Reporting date: 2026-03-05
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-03-05-bullshitbench
Claim: On BullshitBench v2, Anthropic's models excelled at challenging nonsensical prompts rather than confidently answering them while competitors flatlined.
Attributed sources:
- Welcome to March 5, 2026 — https://theinnermostloop.substack.com/p/welcome-to-march-5-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.