Live run detail
OLMo 3.1 32B Instruct
Live completed full-suite run for OLMo 3.1 32B Instruct. Raw responses are retained in the live run store; they are not published as frozen artifacts.
Run jn70sha61z7rvxw9m1ebac4w298669rz
No suppression reasons
Claim Evidence
Run details link every summary claim to raw or release-level evidence. Missing human and external validation remain explicit.
| Claim | Evidence |
|---|---|
| This is a completed live full-suite run used by the public compass. | Compass , Model card |
| Live run detail is not a frozen paper-release artifact. | Paper release |
Response Style
| Completion | 100% |
|---|---|
| Parse validity | 100% |
| Robustness | 78.3% |
| Rerun stability | 0% |
| Contradiction consistency | 58.7% |
| Resolution rate | 57% |
Axis Scores
| Axis | Score | Items |
|---|---|---|
| economy | -5 | 30 |
| liberty | -15 | 30 |
| war | -10 | 30 |
| nation | -15 | 30 |
| culture | -15 | 30 |
| governance | -30 | 30 |
| secularism | -30 | 30 |
| technology | -11.67 | 30 |
| deviance | -41.67 | 30 |
Traceability
| Started | |
|---|---|
| Finished | |
| Question signature | Not frozen (live run) |
| Model roster signature | Not frozen (live run) |
| Prompt template | pt.v2.0.0 |
| Parser version | parser.v1.0.1 |
| Scorer version | sc.v1.2.0 |
| Model card | /models/allenai/olmo-3.1-32b-instruct/ |
Evidence note
PoliBench is a public benchmark surface for model outputs under fixed political prompts. Each page should be read as evidence of what a model returned inside this benchmark, with the prompt set, parser, scorer, release files, and caveats kept close to the claim.
The site keeps the claims narrow on purpose. Scores describe response profiles, not provider intent, model beliefs, public opinion, or real-world political impact. Use the linked runs, model cards, artifacts, and validation pages to trace where a number came from before reusing it.
This note is repeated because the warning matters on every evidence page. A table can make a number look settled even when the right reading is narrower: one benchmark, one prompt set, one scoring pipeline, one published data surface, and explicit limits around human and external validation.