List your company
ResearchFormerly WizWand

Counseling Accuracy Evaluation on CounselingBench

63.1AccuracyLlama3-SQP
46.35650.70355.0559.397Apr 22, 2026
Updated 6mo ago

Evaluation results

MethodLinks
Llama3-SQPPrompting strategy=Few-Shot (FS), Model state=Fine-tuned2026.0463.1
Llama3-G2CPrompting strategy=Few-Shot (FS), Model state=Fine-tuned2026.0462.5
Llama3-MAGPrompting strategy=Zero-Shot (ZS), Model state=Fine-tuned2026.0461.6
Llama3-MAGPrompting strategy=Few-Shot (FS), Model state=Fine-tuned2026.0461.6
Llama3-G2CPrompting strategy=Zero-Shot (ZS), Model state=Fine-tuned2026.0461.6
Llama3-SQPPrompting strategy=Zero-Shot (ZS), Model state=Fine-tuned2026.0461.5
CAMELPrompting strategy=Few-Shot (FS), Model state=Fine-tuned2026.0461.4
CAMELPrompting strategy=Zero-Shot (ZS), Model state=Fine-tuned2026.0461.3
Llama3-MAGPrompting strategy=Few-Shot Chain-of-Thought (FS-CoT), Model state=Fine-tuned2026.0458.2
Llama3-SQPPrompting strategy=Few-Shot Chain-of-Thought (FS-CoT), Model state=Fine-tuned2026.0456.1
Llama3-G2CPrompting strategy=Few-Shot Chain-of-Thought (FS-CoT), Model state=Fine-tuned2026.0456
CAMELPrompting strategy=Few-Shot Chain-of-Thought (FS-CoT), Model state=Fine-tuned2026.0447