# Human Logic Alignment on DailyDialog

Source: [Human Logic Alignment on DailyDialog](<https://www.sota2.com/research/sota/human-logic-alignment-on-dailydialog>)

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

Data updated: 2026-02-26T05:46:42.316Z

Evaluation of human logic alignment performance on the DailyDialog dataset, measured using specific thresholded and non-thresholded alignment scores.

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 | Human Logic Alignment |
| Dataset (including any stated split) | DailyDialog |
| Metric | Human Logic Alignment (T=0.5) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Human Logic Alignment (T=0.5)](<https://www.sota2.com/research/sota/human-logic-alignment-on-dailydialog.md?metric=Human%20Logic%20Alignment%20(T%3D0.5)>): Higher is better; unit: PERCENT.

- [Human Logic Alignment (No Threshold)](<https://www.sota2.com/research/sota/human-logic-alignment-on-dailydialog.md?metric=Human%20Logic%20Alignment%20(No%20Threshold)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Human Logic Alignment (T=0.5) | Result date | Report paper |
| --- | --- | --- | --- |
| MCTS-Driven Knowledge Retrieval | 80.97 | Not recorded | 695b457853fdd5c361743005 |
| MCTS-Driven Knowledge Retrieval | 57.06 | Not recorded | 695b457853fdd5c361743005 |
| ChatGPT Inference | 55.82 | Not recorded | 695b457853fdd5c361743005 |
| Instructor | 45.56 | Not recorded | 695b457853fdd5c361743005 |
| GTE | 43.41 | Not recorded | 695b457853fdd5c361743005 |
| MCTS-Driven Knowledge Retrieval | 42.17 | Not recorded | 695b457853fdd5c361743005 |
| DPR | 42.12 | Not recorded | 695b457853fdd5c361743005 |
| Contriever | 41.1 | Not recorded | 695b457853fdd5c361743005 |
| SBERT | 38.46 | Not recorded | 695b457853fdd5c361743005 |

### MCTS-Driven Knowledge Retrieval

Human Logic Alignment (T=0.5): 80.97. Result ID: 699635e25698665ec4784a24.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

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

### MCTS-Driven Knowledge Retrieval

Human Logic Alignment (T=0.5): 57.06. Result ID: 699635e25698665ec4784a25.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

Recorded evaluation configuration: {"Ablation":"w/o Reasoner"}. This record does not assert independent reproduction.

### ChatGPT Inference

Human Logic Alignment (T=0.5): 55.82. Result ID: 699635e25698665ec4784a23.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bf71e209131cea314c0.md>)

[Implementation](<https://github.com/openai/following-instructions-human-feedback>)

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

### Instructor

Human Logic Alignment (T=0.5): 45.56. Result ID: 699635e25698665ec4784a22.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f901e209131cea46422.md>)

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

### GTE

Human Logic Alignment (T=0.5): 43.41. Result ID: 699635e25698665ec4784a21.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa703b1e209131cea49c20.md>)

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

### MCTS-Driven Knowledge Retrieval

Human Logic Alignment (T=0.5): 42.17. Result ID: 699635e25698665ec4784a26.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

Recorded evaluation configuration: {"Ablation":"w/o Concept Bridging"}. This record does not assert independent reproduction.

### DPR

Human Logic Alignment (T=0.5): 42.12. Result ID: 699635e25698665ec4784a1f.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa686d1e209131cea1c190.md>)

[Implementation](<https://github.com/facebookresearch/DPR>)

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

### Contriever

Human Logic Alignment (T=0.5): 41.1. Result ID: 699635e25698665ec4784a20.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d4b1e209131cea38d50.md>)

[Implementation](<https://github.com/facebookresearch/contriever>)

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

### SBERT

Human Logic Alignment (T=0.5): 38.46. Result ID: 699635e25698665ec4784a1e.

[Reporting paper](<https://www.sota2.com/research/paper/695b457853fdd5c361743005.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66441e209131cea0f5ab.md>)

[Implementation](<https://github.com/UKPLab/sentence-transformers>)

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

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