# Conversational Recommendation on SafeGame

Source: [Conversational Recommendation on SafeGame](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame>)

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

Data updated: 2026-03-05T07:37:32.115Z

Evaluation of Conversational Recommendation performance on the SafeGame dataset, measured using ranking metrics such as Recall@K, NDCG@K, SVR@K, and S-DCG@K.

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 | Conversational Recommendation |
| Dataset (including any stated split) | SafeGame |
| Metric | Recall@5 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 11 of 11 |
| 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

- [Recall@5](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=Recall%405>): Higher is better; unit: PERCENT.

- [Recall@10](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=Recall%4010>): Higher is better; unit: PERCENT.

- [NDCG@5](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=NDCG%405>): Higher is better; unit: PERCENT.

- [NDCG@10](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=NDCG%4010>): Higher is better; unit: PERCENT.

- [SVR@5](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=SVR%405>): Higher is better; unit: PERCENT.

- [SVR@10](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=SVR%4010>): Higher is better; unit: PERCENT.

- [S-DCG@5](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=S-DCG%405>): Higher is better; unit: PERCENT.

- [S-DCG@10](<https://www.sota2.com/research/sota/conversational-recommendation-on-safegame.md?metric=S-DCG%4010>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Recall@5 | Result date | Report paper |
| --- | --- | --- | --- |
| SafeCRS (Llama-3.1-8B) | 27.02 | Not recorded | 69a90de92d603e6de3e40679 |
| SafeCRS (Qwen3-8B) | 23.45 | Not recorded | 69a90de92d603e6de3e40679 |
| SafeCRS (Llama-3.2-3B) | 22.76 | Not recorded | 69a90de92d603e6de3e40679 |
| SafeCRS (Qwen2.5-0.5B) | 15.42 | Not recorded | 69a90de92d603e6de3e40679 |
| Llama3-SDSC-70B | 7.35 | Not recorded | 69a90de92d603e6de3e40679 |
| CRAG (GPT-4o) | 4.22 | Not recorded | 69a90de92d603e6de3e40679 |
| GPT-5.2 | 4.22 | Not recorded | 69a90de92d603e6de3e40679 |
| GPT-4 | 3.63 | Not recorded | 69a90de92d603e6de3e40679 |
| CRAG (Haiku4.5) | 1.62 | Not recorded | 69a90de92d603e6de3e40679 |
| Gemma3-27B | 0.81 | Not recorded | 69a90de92d603e6de3e40679 |
| CRAG (Gemini2.0) | 0.36 | Not recorded | 69a90de92d603e6de3e40679 |

### SafeCRS (Llama-3.1-8B)

Recall@5: 27.02. Result ID: 69a9327f6dbb7075b95241c8.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

Recorded evaluation configuration: {"Backbone":"Llama-3.1-8B"}. This record does not assert independent reproduction.

### SafeCRS (Qwen3-8B)

Recall@5: 23.45. Result ID: 69a9327f6dbb7075b95241c6.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction.

### SafeCRS (Llama-3.2-3B)

Recall@5: 22.76. Result ID: 69a9327f6dbb7075b95241c7.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

Recorded evaluation configuration: {"Backbone":"Llama-3.2-3B"}. This record does not assert independent reproduction.

### SafeCRS (Qwen2.5-0.5B)

Recall@5: 15.42. Result ID: 69a9327f6dbb7075b95241c5.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-0.5B"}. This record does not assert independent reproduction.

### Llama3-SDSC-70B

Recall@5: 7.35. Result ID: 69a9327f6dbb7075b95241c4.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Zero-shot":"true"}. This record does not assert independent reproduction.

### CRAG (GPT-4o)

Recall@5: 4.22. Result ID: 69a9327f6dbb7075b95241be.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77111e209131cea70ef7.md>)

Recorded evaluation configuration: {"Collaborative Retrieval":"true"}. This record does not assert independent reproduction.

### GPT-5.2

Recall@5: 4.22. Result ID: 69a9327f6dbb7075b95241c2.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695f39f053fdd5c361743ecb.md>)

Recorded evaluation configuration: {"Zero-shot":"true"}. This record does not assert independent reproduction.

### GPT-4

Recall@5: 3.63. Result ID: 69a9327f6dbb7075b95241c1.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"Zero-shot":"true"}. This record does not assert independent reproduction.

### CRAG (Haiku4.5)

Recall@5: 1.62. Result ID: 69a9327f6dbb7075b95241bf.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77111e209131cea70ef7.md>)

Recorded evaluation configuration: {"Collaborative Retrieval":"true"}. This record does not assert independent reproduction.

### Gemma3-27B

Recall@5: 0.81. Result ID: 69a9327f6dbb7075b95241c3.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa779f1e209131cea746cd.md>)

Recorded evaluation configuration: {"Zero-shot":"true"}. This record does not assert independent reproduction.

### CRAG (Gemini2.0)

Recall@5: 0.36. Result ID: 69a9327f6dbb7075b95241c0.

[Reporting paper](<https://www.sota2.com/research/paper/69a90de92d603e6de3e40679.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77111e209131cea70ef7.md>)

Recorded evaluation configuration: {"Collaborative Retrieval":"true"}. This record does not assert independent reproduction.

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