Model profile
Llama 3.1 8B Base
2024-07-23release date
1eval lineages
Evidence summary
Published evidence spans 1 evals and 2 of 7 behavior components. Its strongest relative result is AbstentionBench (stale_f1, #16 of 20); its weakest is AbstentionBench (answer_unknown_f1, #20 of 20).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Official model page ↗Family-level model document · meta · first party
- Release source ↗exact first party family release
Published eval results
Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.
| Eval / sub-eval | Rank | Value | Distribution | Better | Source |
|---|---|---|---|---|---|
| AbstentionBenchanswer_unknown_f1 | #20 / 20 | 0.6322 | ↑ higher | Source ↗official | |
| AbstentionBenchfalse_premise_f1 | #19 / 20 | 0.474 | ↑ higher | Source ↗official | |
| AbstentionBenchstale_f1 | #16 / 20 | 0.5877 | ↑ higher | Source ↗official | |
| AbstentionBenchsubjective_f1 | #18 / 20 | 0.5653 | ↑ higher | Source ↗official | |
| AbstentionBenchunderspecified_context_f1 | #16 / 20 | 0.4863 | ↑ higher | Source ↗official | |
| AbstentionBenchunderspecified_intent_f1 | #20 / 20 | 0.5494 | ↑ higher | Source ↗official |