Skip to content
Trending storyDeveloping

Elie Bakouch adds five open models to Blood on the Clocktower benchmark

1 article1 sourcesince Nov 11Last article

Overview

AISummary of 1 article

Elie Bakouch added five open models to Dwarkesh Patel's Blood on the Clocktower benchmark, a social deduction game where models lie, bluff and try to catch each other.

Kimi K3.1 Pro ranked best among the open models, just below GPT-5.5 among closed models.

Bakouch says models that play the game well also identify evil players more accurately when reading only the public transcript. He suggests this could let new models be ranked from existing games without playing new ones.

Written by AI from the articles below · updated Nov 11, 2:43 PM ET

Check the sources:

Latest developmentNov 11, 2:31 PM ET
Kimi K3.1 Pro ranks best open model on Blood on the Clocktower benchmark

Article timeline

The articles in this story. Times are ET.

Nov 11
  1. Dwarkesh PatelX
    Kimi K3.1 Pro ranks best open model on Blood on the Clocktower benchmark

    AIElie Bakouch added five open models to Dwarkesh Patel's Blood on the Clocktower benchmark, and Kimi K3.1 Pro ranked best among them, just below GPT-5.5 among closed models. He says models that do well in the game also spot evil players better when reading only the public transcript, which could allow new models to be ranked from existing games.

Heat Trend

Current comparable heat
6
Peak comparable heat
10Nov 11, 3:00 PM ET
Change in past 24 hours
–

The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.