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

Evidence Grounding on SQuAD v1.1 (test)

80.02F1 Score

AEG (with LFE)

Updated 1mo ago

Evaluation Results

Method	Links
AEG (with LFE) 2026.03		80.02	-	91.16	238
Whisper-Large-v3 + Qwen3-Reranker-8B 2026.03		79.14	3.18	90.11	625
Whisper-Medium + Qwen3-Reranker-8B 2026.03		78.91	3.84	90.9	529
Whisper-Small + Qwen3-Reranker-8B 2026.03		78.29	4.71	90.03	431
Whisper-Large-v3 + bge-reranker-v2-m3 2026.03		74.56	3.18	88.36	599
Whisper-Large-v3 + Qwen3-Reranker-4B 2026.03		74.25	3.18	89.94	614
Whisper-Medium + bge-reranker-v2-m3 2026.03		74.17	3.84	87.96	503
Whisper-Medium + Qwen3-Reranker-4B 2026.03		73.88	3.84	87.32	518
Whisper-Small + bge-reranker-v2-m3 2026.03		73.85	4.71	87.53	405
Whisper-Small + Qwen3-Reranker-4B 2026.03		73.56	4.71	88.97	420