List your company
ResearchFormerly WizWand

Question Answering on ARC-e (test)

96.07AccuracyBLoB
47.325259.980172.63585.2899Oct 2, 2023Mar 16, 2024Aug 30, 2024Feb 13, 2025Jul 29, 2025Jan 12, 2026Jun 28, 2026
Updated 4mo ago

Evaluation results

MethodLinks
BLoBParams (M)=5.4032026.0696.072.350.15
BaRAParams (M)=5.6082026.0696.041.350.12
FVAE-LoRAParams (M)=33.532026.0695.891.740.13
ScalaBLParams (M)=3.7692026.0695.741.920.14
EnsembleParams (M)=11.3052026.0695.712.630.17
MLEParams (M)=3.7682026.0695.643.480.23
LaplaceParams (M)=3.7682026.0695.611.990.15
MAPParams (M)=3.7682026.0695.533.310.22
MC-DropoutParams (M)=3.7682026.0695.513.380.21
C-LoRAParams (M)=4.2752026.0695.511.320.12
LLaMA-2Model Size=70B, Evaluation Protocol=LAT2023.1092.6——
Interaction-Aware Influence Function (Ours)Base Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0590.44——
LESSBase Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0590.11——
RandomBase Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0590.03——
NV-EmbedBase Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0589.01——
Additive IFBase Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0588.72——
LLaMA-2Model Size=70B, Evaluation Protocol=Few-Shot2023.1088.7——
RDS+Base Model=Llama-3.1-8B, Evaluation Protocol=Fine-tuned, Number of Seeds=52026.0588.7——
LLaMA-2Model Size=13B, Evaluation Protocol=LAT2023.1086.3——
LLaMA-2Model Size=13B, Evaluation Protocol=Few-Shot2023.1084.9——
LLAMA-3.1-8BModel Architecture=LLAMA-3.1-8B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0182.53——
QWEN3-14BModel Architecture=QWEN3-14B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0181.9——
MISTRAL-7BModel Architecture=MISTRAL-7B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0181.73——
LLaMA-2Model Size=7B, Evaluation Protocol=LAT2023.1080.3——
LLaMA-2Model Size=7B, Evaluation Protocol=Few-Shot2023.1080.1——
DEEPSEEK-R1-QWEN-8BModel Architecture=DEEPSEEK-R1-QWEN-8B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0176.52——
QWEN3-4BModel Architecture=QWEN3-4B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0175.93——
Pythia-6.9B w/ Qwen2-7BShots=5, Teacher=Qwen2-7B2025.0673.27——
LLAMA-3.2-3BModel Architecture=LLAMA-3.2-3B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0172.05——
Pythia-6.9B w/ Qwen2-7BShots=0, Teacher=Qwen2-7B2025.0671.72——
Pythia-6.9B w/ LLAMA3-8BShots=5, Teacher=LLAMA3-8B2025.0669.78——
Pythia-6.9BShots=52025.0669.23——
Pythia-6.9B w/ Gemma-7BShots=5, Teacher=Gemma-7B2025.0668.94——
Pythia-6.9B w/ Gemma-7BShots=0, Teacher=Gemma-7B2025.0667.59——
Pythia-6.9B w/ LLAMA3-8BShots=0, Teacher=LLAMA3-8B2025.0667.05——
DARTModel Architecture=QWEN3-14B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0166.71——
Pythia-6.9B + Direct tuningShots=02025.0666.25——
Pythia-6.9B + Direct tuningShots=52025.0666.2——
DEEPSEEK-R1-LLAMA-8BModel Architecture=DEEPSEEK-R1-LLAMA-8B, Configuration Type=Dense, Evaluation Protocol=Zero-shot, Sparsity Level=0%2026.0166.2——
Pythia-6.9BShots=02025.0665.99——
Pythia-6.9B + Sentence distillShots=52025.0665.36——
DARTModel Architecture=MISTRAL-7B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0165.07——
Pythia-1B w/ LLAMA3-8BShots=5, Teacher=LLAMA3-8B2025.0664.56——
Pythia-1B w/ LLAMA3-8BShots=0, Teacher=LLAMA3-8B2025.0664.02——
Pythia-1B w/ Qwen2-7BShots=5, Teacher=Qwen2-7B2025.0663.17——
Pythia-1B w/ Qwen2-7BShots=0, Teacher=Qwen2-7B2025.0662.33——
Pythia-6.9B + Sentence distillShots=02025.0661.7——
DARTModel Architecture=DEEPSEEK-R1-QWEN-8B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0161.11——
DARTModel Architecture=LLAMA-3.1-8B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0159.43——
Pythia-1BShots=52025.0658.71——
Pythia-1B + Direct tuningShots=02025.0657.49——
Pythia-1B w/ Gemma-7BShots=5, Teacher=Gemma-7B2025.0656.99——
Pythia-1BShots=02025.0656.82——
Pythia-1B + Direct tuningShots=52025.0655.64——
DARTModel Architecture=DEEPSEEK-R1-LLAMA-8B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0155.43——
Pythia-1B w/ Gemma-7BShots=0, Teacher=Gemma-7B2025.0655.39——
Pythia-1B + Sentence distillShots=52025.0653.41——
Pythia-1B + Sentence distillShots=02025.0652.27——
DARTModel Architecture=LLAMA-3.2-3B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0150.38——
DARTModel Architecture=QWEN3-4B, Configuration Type=Sparse, Evaluation Protocol=Zero-shot, Sparsity Level=70%2026.0149.2——