Commonsense Reasoning on StrategyQA (Accuracy (%))

76.13Accuracy (%)

Dream 7B-Instruct (Top Margin + Suffix-Anchored Confidence Modulation)

Updated 2d ago

Evaluation Results

Method	Links
Dream 7B-Instruct (Top Margin + Suffix-Anchored Confidence Modulation) 2026.05		76.13
Dream 7B-Instruct (Top Margin + Suffix Anchor) 2026.05		75.98
Dream 7B-Instruct (Top Probability + Suffix-Anchored Confidence Modulation) 2026.05		75.4
Dream 7B-Instruct (Top Probability + Suffix Anchor) 2026.05		74.24
Vanilla 2026.05		73.6
COFT 2026.05		73.6
DT-CD⋆ 2026.05		73.5
SDD 2026.05		73
DExperts 2026.05		72.8
Vanilla 2026.05		71.2
COFT 2026.05		71.1
DT-CD⋆ 2026.05		71
SDD 2026.05		70.5
LLaDA 8B-Instruct (Top Probability + Suffix-Anchored Confidence Modulation) 2026.05		70.45
DExperts 2026.05		70.3
Dream 7B-Instruct (Random) 2026.05		69.43
LLaDA 8B-Instruct (Top Margin + Suffix-Anchored Confidence Modulation) 2026.05		68.12
LLaDA 8B-Instruct (Top Probability + Suffix Anchor) 2026.05		66.52
LLaDA 8B-Instruct (Top Margin + Suffix Anchor) 2026.05		66.08
Dream 7B-Instruct (Top Margin) 2026.05		62.45
LLaDA 8B-Instruct (Random) 2026.05		58.81
Dream 7B-Instruct (Top Probability) 2026.05		56.91
LLaDA 8B-Instruct (Top Margin) 2026.05		28.53
LLaDA 8B-Instruct (Top Probability) 2026.05		19.65