Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

Short-form QA on SimpleQA

4Accuracy

CRL-Token

Updated 1mo ago

Evaluation Results

Method	Links
CRL-Token 2026.02		4	0.22
Gemma 2 2B 2026.02		3.78	-

SOTA Paper

CRL-Token

Control Reinforcement Learning: Interpretable Token-Level Steering of LLMs via Sparse Autoencoder Features

Dataset

SimpleQA

Follow for update

@wizwand_team Discord

Related Benchmarks

Bias QA on BBQ Ambig Bias QA on BBQ Disambig Math Reasoning on GSM8K (Accuracy and Gain)Adversarial Safety on HarmBench Multiple Choice Question Answering on MMLU (Performance Change Tracking)

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs