List your company

Ambiguity Detection on AmbigQA

79.5F1 ScoreH0 topology
32.1844.46556.7569.035May 27, 2026May 31, 2026Jun 4, 2026Jun 9, 2026Jun 13, 2026Jun 17, 2026Jun 22, 2026

Evaluation results

MethodLinks
H0 topologyModel=Gemma-7B-it2026.0679.5——79.5
H0 topologyModel=Llama-3.1-8B-Instruct2026.0679.3——79.3
H0 topologyModel=Mistral-7B-Instruct-v0.32026.0677.8——77.9
AENModel=Mistral-7B-Instruct-v0.32026.0677.3——77.3
ShaQ (Max Span)Backbone=GPT-42026.0570.3864.259.95—
ShaQBackbone=GPT-42026.0570.3865.7761.52—
LOOBackbone=GPT-42026.0567.7750.4354.35—
LOO (Max Span)Backbone=GPT-42026.0567.7752.3655.8—
ICEBackbone=GPT-42026.0565.5460.6659.98—
AENModel=Llama-3.1-8B-Instruct2026.0663.2——63.4
AENModel=Gemma-7B-it2026.0661.6——61.6
InfoGainModel=Mistral-7B-Instruct-v0.32026.0659.8——59.8
Deep EnsemblesBackbone=GPT-42026.0559.760.2960.31—
Few-shotModel=Mistral-7B-Instruct-v0.32026.0658.4——58.4
Sample RepetitionBackbone=GPT-42026.0556.9860.959.24—
Ask4Conf-DBackbone=GPT-42026.0556.3652.4453.97—
Sample DiversityBackbone=GPT-42026.0555.2460.8659.35—
Self-ConsistencyBackbone=GPT-42026.0553.459.3357.59—
InfoGainModel=Gemma-7B-it2026.0650.7——52.5
Zero-shotModel=Llama-3.1-8B-Instruct2026.0645.9——50
Few-shot CoTModel=Llama-3.1-8B-Instruct2026.0645.8——49.1
Semantic EntropyBackbone=GPT-42026.0544.5857.1657—
Few-shot CoTModel=Mistral-7B-Instruct-v0.32026.0643.1——49.2
Few-shot CoTModel=Gemma-7B-it2026.0642.7——42.7
Zero-shotModel=Gemma-7B-it2026.0641.8——50.4
InfoGainModel=Llama-3.1-8B-Instruct2026.0640.9——52.9
Few-shotModel=Gemma-7B-it2026.0636.3——50.4
Few-shotModel=Llama-3.1-8B-Instruct2026.0635.5——52.5
Zero-shotModel=Mistral-7B-Instruct-v0.32026.0634——49.4