# SCONE post-cutoff success leaderboard

external benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/external-scone-post-cutoff-success

SCONE Post-Cutoff Exploit Success. Every model the catalog carries a published SCONE post-cutoff success value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | external |
| Measure | Best@8 exploit success rate |
| Tasks | 12 post-cutoff smart-contract vulnerabilities |
| Difficulty | Smart-contract exploitation |
| Published by | [Evaluating frontier models on software exploitation](https://www.anthropic.com/research/exploit-evals) |

Share of SCONE smart-contract vulnerabilities exploited on the 12-task post-cutoff set.

## SCONE post-cutoff success ranking

1 model with a published SCONE post-cutoff success value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Best@8 exploit success rate | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Claude Mythos Preview](https://modelscale.dev/models/claude-mythos-preview) | Anthropic | 100 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
