LoFa - BoolQ: leaderboard

Metric: Logical Fallacy Resistance after three attack rounds (LFR@3, %): among the questions a model first answers correctly and whose plain false claim it rejects, the share it still answers correctly after up to three rounds of a DeepSeek-R1-written argument using one of ten logical fallacies, mean over the ten fallacy types, three runs; higher is better; 491 BoolQ yes or no items. Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.

Top models

#ModelScore
1GPT-491.08
2DeepSeek V379.17
3Llama 3.1 405B73.61
4GPT-3.5 Turbo66.01
5Llama 3.1 70B64.7
6Llama 3.1 8B62.14

Interactive version: theaggregate.ai/benchmark?slug=lofa-boolq · How It Works · Data refreshed daily, snapshot 2026-09-29.