Model profile
Qwen 2.5 0.5B Instruct
Alibabadeveloper
2024-09-19release date
#307 / 309overall rank
4eval lineages
Evidence summary
Qwen 2.5 0.5B Instruct has an estimated overall rank of #307; its 90% source-sensitivity interval is #281–#309. Its behavior-only rank is #306; company governance moves the combined estimate to #307. Published evidence spans 4 evals and 3 of 7 behavior components. Its strongest relative result is Open LLM Safety Index (jailbreakbench_safety_rate, #12 of 21); its weakest is M3-SafetyBench (overall_score, #19 of 19).
Compare this model
Only models sharing at least one published sub-eval are listed.
Official and reference links
- Hugging Face ↗Qwen/Qwen2.5-0.5B-Instruct
- Official model page ↗Exact model document · Reviewed official Hugging Face owner · official repository
- Release source ↗direct research preview identity date
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 | Source |
|---|---|---|---|---|
| Enkrypt AI Safety Leaderboardbias_attack_non_success_rate | #234 / 241 | ↑3.88 | Source ↗official | |
| Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate | #192 / 241 | ↑81.33 | Source ↗official | |
| Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate | #229 / 241 | ↑31.11 | Source ↗official | |
| Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate | #172 / 239 | ↑93.36 | Source ↗official | |
| M3-SafetyBenchoverall_score | #19 / 19 | ↑69.63 | Source ↗official | |
| Open LLM Safety Indexjailbreakbench_safety_rate | #12 / 21 | ↑0.3333 | Source ↗official | |
| Open LLM Safety Indexstrongreject_safety_rate | #16 / 21 | ↑0.0667 | Source ↗official | |
| PandaBench JBB direct-request panelsafety_rate | #40 / 46 | ↑0.94 | Source ↗official |