# Long-context Question Answering on HotpotQA Fixed Chunk 2048 Source: [Long-context Question Answering on HotpotQA Fixed Chunk 2048]() Collection index: [llms.txt]() Data updated: 2026-03-06T09:56:50.180Z Evaluation of model performance on the HotpotQA dataset using a fixed context chunk size of 2048 tokens for long-context question answering tasks. 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) | HotpotQA Fixed Chunk 2048 | | Metric | QA Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 - [QA Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | QA Score | Result date | Report paper | | --- | --- | --- | --- | | Baseline | 60.03 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our | 59.67 | Not recorded | 69aa5f972d603e6de3e41a04 | | Baseline | 59.22 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our + Reorder | 58.2 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our | 57.39 | Not recorded | 69aa5f972d603e6de3e41a04 | | Baseline | 54.1 | Not recorded | 69aa5f972d603e6de3e41a04 | | EPIC (15%) | 53.62 | Not recorded | 69aa5f972d603e6de3e41a04 | | CacheBlend | 53.52 | Not recorded | 69aa5f972d603e6de3e41a04 | | EPIC (15%) | 52.84 | Not recorded | 69aa5f972d603e6de3e41a04 | | CacheBlend | 51.77 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our | 51.5 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our + Reorder | 50.53 | Not recorded | 69aa5f972d603e6de3e41a04 | | Our + Reorder | 50.53 | Not recorded | 69aa5f972d603e6de3e41a04 | | No Recompute | 50.24 | Not recorded | 69aa5f972d603e6de3e41a04 | | EPIC (15%) | 47.55 | Not recorded | 69aa5f972d603e6de3e41a04 | | CacheBlend | 47.2 | Not recorded | 69aa5f972d603e6de3e41a04 | | No Recompute | 46.71 | Not recorded | 69aa5f972d603e6de3e41a04 | | No Recompute | 46.33 | Not recorded | 69aa5f972d603e6de3e41a04 | ### Baseline QA Score: 60.03. Result ID: 69aaa4d76dbb7075b954d391. [Reporting paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our QA Score: 59.67. Result ID: 69aaa4d76dbb7075b954d387. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Baseline QA Score: 59.22. Result ID: 69aaa4d76dbb7075b954d385. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our + Reorder QA Score: 58.2. Result ID: 69aaa4d76dbb7075b954d394. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our QA Score: 57.39. Result ID: 69aaa4d76dbb7075b954d393. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Baseline QA Score: 54.1. Result ID: 69aaa4d76dbb7075b954d38b. [Reporting paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### EPIC (15%) QA Score: 53.62. Result ID: 69aaa4d76dbb7075b954d396. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### CacheBlend QA Score: 53.52. Result ID: 69aaa4d76dbb7075b954d389. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### EPIC (15%) QA Score: 52.84. Result ID: 69aaa4d76dbb7075b954d38a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### CacheBlend QA Score: 51.77. Result ID: 69aaa4d76dbb7075b954d395. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our QA Score: 51.5. Result ID: 69aaa4d76dbb7075b954d38d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our + Reorder QA Score: 50.53. Result ID: 69aaa4d76dbb7075b954d388. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### Our + Reorder QA Score: 50.53. Result ID: 69aaa4d76dbb7075b954d38e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### No Recompute QA Score: 50.24. Result ID: 69aaa4d76dbb7075b954d392. [Reporting paper]() Recorded evaluation configuration: {"Model":"ChatGLM","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### EPIC (15%) QA Score: 47.55. Result ID: 69aaa4d76dbb7075b954d390. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### CacheBlend QA Score: 47.2. Result ID: 69aaa4d76dbb7075b954d38f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### No Recompute QA Score: 46.71. Result ID: 69aaa4d76dbb7075b954d38c. [Reporting paper]() Recorded evaluation configuration: {"Model":"LLaMA","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. ### No Recompute QA Score: 46.33. Result ID: 69aaa4d76dbb7075b954d386. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen","Setting":"Fixed Chunk (2048)"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.