# Long-context Question Answering on Qasper (Extract/Abstract/Bool F1) Source: [Long-context Question Answering on Qasper (Extract/Abstract/Bool F1)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:42:38.563Z Evaluation of model performance on the Qasper dataset for Long-context Question Answering, specifically reporting Extract F1, Abstract F1, and Bool F1 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 | Long-context Question Answering | | Dataset (including any stated split) | Qasper | | Metric | Extract F1 | | 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 - [Extract F1](): Higher is better; unit: PERCENT. - [Abstract F1](): Higher is better; unit: PERCENT. - [Bool F1](): Higher is better; unit: PERCENT. ## Recorded results | Method | Extract F1 | Result date | Report paper | | --- | --- | --- | --- | | GLM-4.1V-9B-Thinking VERA | 54.57 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | GLM-4.1V-9B-Thinking ColPali RAG | 47.51 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | GLM-4.1V-9B-Thinking Random RAG | 46.92 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | GLM-4.1V-9B-Thinking Embedding RAG | 45.21 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | GLM-4.1V-9B-Thinking | 41.69 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | Qwen3-VL-8B-Instruct VERA | 39.29 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | Glyph | 39.26 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | Qwen3-VL-8B-Instruct | 25.54 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | Qwen3-VL-8B-Instruct OCR RAG | 25.15 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | | Qwen3-VL-8B-Instruct Random RAG | 23.96 | Not recorded | 698d5e64c97a0b3e3b5e3c66 | ### GLM-4.1V-9B-Thinking VERA Extract F1: 54.57. Result ID: 699723a25698665ec471f691. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"Attention-Guided RAG"}. This record does not assert independent reproduction. ### GLM-4.1V-9B-Thinking ColPali RAG Extract F1: 47.51. Result ID: 699723a25698665ec471f68e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"ColPali RAG"}. This record does not assert independent reproduction. ### GLM-4.1V-9B-Thinking Random RAG Extract F1: 46.92. Result ID: 699723a25698665ec471f68c. [Reporting paper]() Recorded evaluation configuration: {"RAG Strategy":"Random RAG"}. This record does not assert independent reproduction. ### GLM-4.1V-9B-Thinking Embedding RAG Extract F1: 45.21. Result ID: 699723a25698665ec471f68d. [Reporting paper]() Recorded evaluation configuration: {"RAG Strategy":"Embedding RAG"}. This record does not assert independent reproduction. ### GLM-4.1V-9B-Thinking Extract F1: 41.69. Result ID: 699723a25698665ec471f68b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"Direct (None)"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct VERA Extract F1: 39.29. Result ID: 699723a25698665ec471f690. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"Attention-Guided RAG"}. This record does not assert independent reproduction. ### Glyph Extract F1: 39.26. Result ID: 699723a25698665ec471f68f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"Direct (None)"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct Extract F1: 25.54. Result ID: 699723a25698665ec471f688. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"Direct (None)"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct OCR RAG Extract F1: 25.15. Result ID: 699723a25698665ec471f68a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"RAG Strategy":"OCR RAG"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct Random RAG Extract F1: 23.96. Result ID: 699723a25698665ec471f689. [Reporting paper]() Recorded evaluation configuration: {"RAG Strategy":"Random RAG"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.