AI Paper Atlas

Research map / Large language models / Agents, code and math

CounterBench: Evaluating and Improving Counterfactual Reasoning in Large Language Models

Yuefei Chen, Vivek K. Singh, Jing Ma, Ruixiang Tang

AAAI 2026 · Large language models › Agents, code and math · approach: Mathematical and structural reasoning benchmarks

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use CoIn reasoning instead of CausalCoTGPT-4o · CounterBenchaccuracy: 75.8 -> 89.4Improved
use CoIn reasoning instead of CausalCoTGPT-4o mini · CounterBenchaccuracy: 59.8 -> 79.9Improved
use CoIn reasoning instead of CausalCoTGPT-3 (Babbage-002) and Davinci-002 · CounterBenchaccuracy: 49.3 -> 49.6 (Davinci-002)No clear effect
add CausalCoT prompting over standard promptingmost LLMs · CounterBenchaccuracy: 52.8 -> 75.8 (GPT-4o); 50.6 -> 49.1 (GPT-3.5)Mixed
add tool-based Solver strategy (CausalTool re-implementation)LLMs · CounterBenchaccuracy: 50.4 -> 47.2 (GPT-4o mini)Hurt
apply CoIn to Gemini-1.5-flash instead of standard promptingGemini-1.5-flash · CounterBenchaccuracy: 68.0 -> 89.9Improved
apply CoIn on CLADDER counterfactual subsetGemini-1.5 · CLADDER counterfactualaccuracy (%): 64.77 -> 78.98Improved
switch CLADDER queries from commonsense to anti-commonsense (ablation)CoIn · CLADDER counterfactualaccuracyNo clear effect
increase number of events (variables) per question (ablation)CausalCoT and CoIn · CounterBenchaccuracyHurt
apply CoIn to small models (GPT-4o mini, Claude-3.5 Haiku, Gemini-1.5-flash-8b)small LLMs · CounterBenchaccuracy: 50.4 -> 79.9 (GPT-4o mini, standard vs CoIn)Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy