| NLP Task Suite (Capitalize, Country-Capital, Present-Past, Singular-Plural, Person-Sport, AG News) (test) | ICL baseline | Capitalize99.9 | | 20 | 4mo ago |
| 12 Downstream Classification Tasks | StateX | Accuracy53 | | 15 | 3mo ago |
| 9-dataset Average (SST-5, MNLI, CMSQA, HellaSwag, GeoQ, NL2Bash, Break, MTOP, SMCalFlow) (test) | CLG | Accuracy68.08 | | 15 | 4mo ago |
| Garg ICL d=20 2022 (held-out tasks) | Transformer (continuous attn) | R20.989 | | 10 | 22d ago |
| Garg ICL (d=10) (held-out tasks) | Transformer (continuous attn) | R20.996 | | 9 | 22d ago |
| Fineweb-Edu 16.8B tokens | Spectra-AdEMAMix | ARC-c Accuracy36.86 | | 8 | 4mo ago |
| ChemBench | ABMLL-MetaICL | Accuracy58.4 | | 6 | 3mo ago |
| LegalBench | ABMLL-MetaICL | Accuracy79.5 | | 6 | 3mo ago |
| Llama3-8B Scenario 5 ICL prompts | LTV | Accuracy82.8 | | 3 | 3mo ago |
| Llama3-8B Scenario 4: More layers & Pos. (P={-5,...}, L={0,4,...}) | LTV | Accuracy46.38 | | 3 | 3mo ago |
| Llama3-8B Scenario 3: More layers (L={0,4,8,...}) | LTV | Accuracy80.43 | | 3 | 3mo ago |
| Llama3-8B Scenario 2: More Pos. (P={-5,...,-1}) | LTV | Accuracy78.18 | | 3 | 3mo ago |
| Llama3-8B Scenario 1: Diff. Pos. (P={4}) | LTV | Accuracy74.1 | | 3 | 3mo ago |
| Llama3-8B Baseline (P={-1}, L={14}) | LTV | Accuracy78.65 | | 3 | 3mo ago |
| MMLongBench iNat | FOCUS | Accuracy22 | | 2 | 1mo ago |
| MMLongBench Food | FOCUS | Accuracy34 | | 2 | 1mo ago |
| MMLongBench Cars | FOCUS | Accuracy0.2642 | | 2 | 1mo ago |
