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

Home/Benchmarks

Conversational Behavior Reasoning on CANDOR

0.58BLEU-1

GoT

Updated 1mo ago

Evaluation Results

Method	Links
GoT 2025.12		0.58	0.56	0.49	0.66

SOTA Paper

GoT

Enabling Conversational Behavior Reasoning Capabilities in Full-Duplex Speech

Dataset

CANDOR

Follow for update

@wizwand_team Discord

Related Benchmarks

Conversational Behavior Detection on CANDOR Rationale Generation on GPT-4 full-duplex audio Rationale Generation on Moshi full-duplex Rationale Generation on Simulation dataset audio Conversational Behavior Reasoning on Synthetic

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs