List your company

Question Answering on GPQA Diamond (solve rate)

44.44Solve RateBifrost
-0.727210.998922.72534.4511Feb 5, 2026

Evaluation results

MethodLinks
BifrostBackbone=Llama-3.2 1B2026.0244.44
BifrostBackbone=Llama-3.1 8B2026.0237.88
ReflexionBackbone=Llama-3.1 8B2026.0233.84
BifrostBackbone=Llama-3.2 3B2026.0232.83
DoTBackbone=Llama-3.1 8B2026.0230.81
ICLBackbone=Llama-3.2 3B2026.0225.25
ICLBackbone=Llama-3.2 1B2026.0223.23
ICLBackbone=Llama-3.1 8B2026.0222.22
RISEBackbone=Llama-3.2 1B, Attempts=52026.0221.21
ReflexionBackbone=Llama-3.2 3B2026.0220.2
PaprikaBackbone=Llama-3.2 3B2026.0220.2
RISEBackbone=Llama-3.2 3B, Attempts=52026.0219.19
PaprikaBackbone=Llama-3.2 1B2026.0218.69
CoTBackbone=Llama-3.1 8B2026.0217.68
BoTBackbone=Llama-3.2 1B2026.0216.67
DoTBackbone=Llama-3.2 3B2026.0213.64
CoTBackbone=Llama-3.2 3B2026.0212.63
RISEBackbone=Llama-3.1 8B, Attempts=52026.0212.63
BoTBackbone=Llama-3.2 3B2026.0212.12
PaprikaBackbone=Llama-3.1 8B2026.0212.12
BoTBackbone=Llama-3.1 8B2026.0211.11
RISEBackbone=Llama-3.2 1B, Attempts=12026.026.06
CoTBackbone=Llama-3.2 1B2026.024.04
RISEBackbone=Llama-3.2 3B, Attempts=12026.024.04
RISEBackbone=Llama-3.1 8B, Attempts=12026.023.54
ReflexionBackbone=Llama-3.2 1B2026.021.1
DoTBackbone=Llama-3.2 1B2026.021.01