# Privacy-Aware Minimal View for Agents on Augmented WebArena N = 662 (test)

Source: [Privacy-Aware Minimal View for Agents on Augmented WebArena N = 662 (test)](<https://www.sota2.com/research/sota/privacy-aware-minimal-view-for-agents-on-augmented-webarena-n-662-test>)

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

Data updated: 2026-06-15T07:02:13.392Z

Evaluation of agent performance focusing on privacy-aware minimal view settings on the Augmented WebArena N=662 test set, measured by Context Recall (TCNP), TCNP-I, TISL, and Keep%.

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 | Privacy-Aware Minimal View for Agents |
| Dataset (including any stated split) | Augmented WebArena N = 662 (test) |
| Metric | Context Recall (TCNP) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [Context Recall (TCNP)](<https://www.sota2.com/research/sota/privacy-aware-minimal-view-for-agents-on-augmented-webarena-n-662-test.md?metric=Context%20Recall%20(TCNP)>): Higher is better; unit: PERCENT.

- [TCNP-I](<https://www.sota2.com/research/sota/privacy-aware-minimal-view-for-agents-on-augmented-webarena-n-662-test.md?metric=TCNP-I>): Higher is better; unit: PERCENT.

- [TISL](<https://www.sota2.com/research/sota/privacy-aware-minimal-view-for-agents-on-augmented-webarena-n-662-test.md?metric=TISL>): Higher is better; unit: PERCENT.

- [Keep%](<https://www.sota2.com/research/sota/privacy-aware-minimal-view-for-agents-on-augmented-webarena-n-662-test.md?metric=Keep%25>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Context Recall (TCNP) | Result date | Report paper |
| --- | --- | --- | --- |
| Llama-3.3-70B-Instruct | 96.6 | Not recorded | 6a2f871cd33dfb23e783c014 |
| GPT-OSS-20B | 96.3 | Not recorded | 6a2f871cd33dfb23e783c014 |
| Nemotron-Nano-9B | 95.8 | Not recorded | 6a2f871cd33dfb23e783c014 |
| Qwen3-8B-Instruct | 95.6 | Not recorded | 6a2f871cd33dfb23e783c014 |
| Llama-3-8B-Instruct | 95.5 | Not recorded | 6a2f871cd33dfb23e783c014 |
| Mistral-7B-v0.3 | 95.3 | Not recorded | 6a2f871cd33dfb23e783c014 |
| Gemma-3N-E4B | 95.1 | Not recorded | 6a2f871cd33dfb23e783c014 |
| MINIM | 94.91 | Not recorded | 6a2f871cd33dfb23e783c014 |

### Llama-3.3-70B-Instruct

Context Recall (TCNP): 96.6. Result ID: 6a2fa35ffbc373592c2b6672.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### GPT-OSS-20B

Context Recall (TCNP): 96.3. Result ID: 6a2fa35ffbc373592c2b6671.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### Nemotron-Nano-9B

Context Recall (TCNP): 95.8. Result ID: 6a2fa35ffbc373592c2b6670.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### Qwen3-8B-Instruct

Context Recall (TCNP): 95.6. Result ID: 6a2fa35ffbc373592c2b666f.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### Llama-3-8B-Instruct

Context Recall (TCNP): 95.5. Result ID: 6a2fa35ffbc373592c2b6674.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### Mistral-7B-v0.3

Context Recall (TCNP): 95.3. Result ID: 6a2fa35ffbc373592c2b6673.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### Gemma-3N-E4B

Context Recall (TCNP): 95.1. Result ID: 6a2fa35ffbc373592c2b6675.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Zero-shot","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

### MINIM

Context Recall (TCNP): 94.91. Result ID: 6a2fa35ffbc373592c2b6676.

[Reporting paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2f871cd33dfb23e783c014.md>)

Recorded evaluation configuration: {"Protocol":"Proposed specialized approach","Input":"Text-only accessibility trees"}. This record does not assert independent reproduction.

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