# Preference-aligned decision making on AmbiK (test)

Source: [Preference-aligned decision making on AmbiK (test)](<https://www.sota2.com/research/sota/preference-aligned-decision-making-on-ambik-test>)

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

Data updated: 2026-05-14T07:05:05.872Z

Evaluation of model performance on Preference-aligned decision making using the test split of the AmbiK dataset, measured by 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 | Preference-aligned decision making |
| Dataset (including any stated split) | AmbiK (test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/preference-aligned-decision-making-on-ambik-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Adaptive CLIPR | 84.6 | Not recorded | 6a055716dbc836b87d828cac |
| CLIPR | 82.6 | Not recorded | 6a055716dbc836b87d828cac |
| ICL + Answers | 82.4 | Not recorded | 6a055716dbc836b87d828cac |
| TidyBot | 78.6 | Not recorded | 6a055716dbc836b87d828cac |
| CIPHER (Sem.) | 71.7 | Not recorded | 6a055716dbc836b87d828cac |
| CIPHER (Lev.) | 69.4 | Not recorded | 6a055716dbc836b87d828cac |
| GATE (15-turn) | 62.3 | Not recorded | 6a055716dbc836b87d828cac |
| ICL | 53.8 | Not recorded | 6a055716dbc836b87d828cac |
| IP | 51.7 | Not recorded | 6a055716dbc836b87d828cac |
| Zero-shot | 50.2 | Not recorded | 6a055716dbc836b87d828cac |

### Adaptive CLIPR

Accuracy: 84.6. Result ID: 6a05734647f34ce924ebeb77.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

Recorded evaluation configuration: {"Feedback":"Adaptive"}. This record does not assert independent reproduction.

### CLIPR

Accuracy: 82.6. Result ID: 6a05734647f34ce924ebeb78.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

Recorded evaluation configuration: {"Feedback":"Iterative learning approach"}. This record does not assert independent reproduction.

### ICL + Answers

Accuracy: 82.4. Result ID: 6a05734647f34ce924ebeb7a.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

Recorded evaluation configuration: {"Mode":"In-context learning with answers"}. This record does not assert independent reproduction.

### TidyBot

Accuracy: 78.6. Result ID: 6a05734647f34ce924ebeb79.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa713f1e209131cea5013a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### CIPHER (Sem.)

Accuracy: 71.7. Result ID: 6a05734647f34ce924ebeb7c.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69fde.md>)

[Implementation](<https://github.com/gao-g/prelude>)

Recorded evaluation configuration: {"Distance Metric":"Semantic"}. This record does not assert independent reproduction.

### CIPHER (Lev.)

Accuracy: 69.4. Result ID: 6a05734647f34ce924ebeb7b.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69fde.md>)

[Implementation](<https://github.com/gao-g/prelude>)

Recorded evaluation configuration: {"Distance Metric":"Levenshtein"}. This record does not assert independent reproduction.

### GATE (15-turn)

Accuracy: 62.3. Result ID: 6a05734647f34ce924ebeb7d.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70e51e209131cea4dc31.md>)

Recorded evaluation configuration: {"Interaction turns":"15"}. This record does not assert independent reproduction.

### ICL

Accuracy: 53.8. Result ID: 6a05734647f34ce924ebeb7e.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa751c1e209131cea65da7.md>)

Recorded evaluation configuration: {"Mode":"In-context learning"}. This record does not assert independent reproduction.

### IP

Accuracy: 51.7. Result ID: 6a05734647f34ce924ebeb7f.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76e41e209131cea6fb66.md>)

Recorded evaluation configuration: {"Method Identity":"Introspective Planning"}. This record does not assert independent reproduction.

### Zero-shot

Accuracy: 50.2. Result ID: 6a05734647f34ce924ebeb80.

[Reporting paper](<https://www.sota2.com/research/paper/6a055716dbc836b87d828cac.md>)

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

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