# Listener Task on PhotoBook (val) Source: [Listener Task on PhotoBook (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:28:59.122Z Evaluation of Listener Task performance on the validation split of the PhotoBook dataset, measured using 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 | Listener Task | | Dataset (including any stated split) | PhotoBook (val) | | Metric | Accuracy | | 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 - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Human | 95 | Not recorded | 68fa6fc91e209131cea47623 | | DeBERTa-based listener model (Ours) | 84.8 | Not recorded | 68fa6fc91e209131cea47623 | | DeBERTa-based listener model (Ours + VisAttn) | 75 | Not recorded | 68fa6fc91e209131cea47623 | | DeBERTa-based listener model (Ours - CLIPScore) | 70.7 | Not recorded | 68fa6fc91e209131cea47623 | | DeBERTa-based listener model (Ours - CLIPScore + VisAttn) | 69.8 | Not recorded | 68fa6fc91e209131cea47623 | | Modified (Takmaz et al., 2020) w/ CLIPScore ref chains | 65 | Not recorded | 68fa6fc91e209131cea47623 | | Modified (Takmaz et al., 2020) | 64.2 | Not recorded | 68fa6fc91e209131cea47623 | | DeBERTa-based listener model (Ours - Dense learning signals) | 59.4 | Not recorded | 68fa6fc91e209131cea47623 | | Random guess | 50 | Not recorded | 68fa6fc91e209131cea47623 | ### Human Accuracy: 95. Result ID: 6990ebb83a6c314849044e14. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DeBERTa-based listener model (Ours) Accuracy: 84.8. Result ID: 6990ebb83a6c314849044e0f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore":true,"Dense learning signals":true,"VisAttn":false}. This record does not assert independent reproduction. ### DeBERTa-based listener model (Ours + VisAttn) Accuracy: 75. Result ID: 6990ebb83a6c314849044e10. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore":true,"Dense learning signals":true,"VisAttn":true}. This record does not assert independent reproduction. ### DeBERTa-based listener model (Ours - CLIPScore) Accuracy: 70.7. Result ID: 6990ebb83a6c314849044e11. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore":false,"Dense learning signals":true,"VisAttn":false}. This record does not assert independent reproduction. ### DeBERTa-based listener model (Ours - CLIPScore + VisAttn) Accuracy: 69.8. Result ID: 6990ebb83a6c314849044e12. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore":false,"Dense learning signals":true,"VisAttn":true}. This record does not assert independent reproduction. ### Modified (Takmaz et al., 2020) w/ CLIPScore ref chains Accuracy: 65. Result ID: 6990ebb83a6c314849044e0e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore ref chains":true}. This record does not assert independent reproduction. ### Modified (Takmaz et al., 2020) Accuracy: 64.2. Result ID: 6990ebb83a6c314849044e0d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"seed runs":3}. This record does not assert independent reproduction. ### DeBERTa-based listener model (Ours - Dense learning signals) Accuracy: 59.4. Result ID: 6990ebb83a6c314849044e13. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"seed runs":3,"CLIPScore":true,"Dense learning signals":false,"VisAttn":false}. This record does not assert independent reproduction. ### Random guess Accuracy: 50. Result ID: 6990ebb83a6c314849044e0c. [Reporting paper]() 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.