# Agentic Long-context Reasoning on CL-bench (test)

Source: [Agentic Long-context Reasoning on CL-bench (test)](<https://www.sota2.com/research/sota/agentic-long-context-reasoning-on-cl-bench-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-20T09:25:49.971Z

Evaluation of model performance on Agentic Long-context Reasoning using the test split of the CL-bench dataset, measured by Solve rate and Rubric accuracy.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Agentic Long-context Reasoning |
| Dataset (including any stated split) | CL-bench (test) |
| Metric | Solve Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Solve Rate](<https://www.sota2.com/research/sota/agentic-long-context-reasoning-on-cl-bench-test.md?metric=Solve%20Rate>): Higher is better; unit: PERCENT.

- [Rubric Accuracy](<https://www.sota2.com/research/sota/agentic-long-context-reasoning-on-cl-bench-test.md?metric=Rubric%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Solve Rate | Result date | Report paper |
| --- | --- | --- | --- |
| RLM + PEEK | 26 | Not recorded | 6a0d402ff984739ed6737955 |
| RLM + Compaction Agent | 20 | Not recorded | 6a0d402ff984739ed6737955 |
| RLM + ACE (Online Adaptation) | 20 | Not recorded | 6a0d402ff984739ed6737955 |
| RLM | 14 | Not recorded | 6a0d402ff984739ed6737955 |
| RLM + RAG | 14 | Not recorded | 6a0d402ff984739ed6737955 |
| RLM + Shared Chat | 12 | Not recorded | 6a0d402ff984739ed6737955 |

### RLM + PEEK

Solve Rate: 26. Result ID: 6a0d7e0bfbc373592c05d8df.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Implementation](<https://github.com/zhuohangu/peek>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

### RLM + Compaction Agent

Solve Rate: 20. Result ID: 6a0d7e0bfbc373592c05d8dd.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79c71e209131cea7f569.md>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

### RLM + ACE (Online Adaptation)

Solve Rate: 20. Result ID: 6a0d7e0bfbc373592c05d8de.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be01e209131cea8afd2.md>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

### RLM

Solve Rate: 14. Result ID: 6a0d7e0bfbc373592c05d8da.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6955ff9153fdd5c36173e066.md>)

[Implementation](<https://github.com/alexzhang13/rlm/tree/main>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

### RLM + RAG

Solve Rate: 14. Result ID: 6a0d7e0bfbc373592c05d8dc.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69b91e209131cea23c26.md>)

[Implementation](<https://github.com/huggingface/transformers/blob/master/examples/rag/>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

### RLM + Shared Chat

Solve Rate: 12. Result ID: 6a0d7e0bfbc373592c05d8db.

[Reporting paper](<https://www.sota2.com/research/paper/6a0d402ff984739ed6737955.md>)

Recorded evaluation configuration: {"Base LM":"GPT-5-mini-2025-08-07"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.