Short-context Reading Comprehension on SQuAD
86.35F1 ScoreBase+ctx
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Base+ctxLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 86.35 | 86.2 | |
| Base+ctxLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 78.15 | 77.89 | |
| D2LckptLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 73.2 | 72.96 | |
| Doc2AtomLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 59.82 | 59.66 | |
| Doc2AtomLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 59.71 | 59.71 | |
| D2LckptLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 29.6 | 29.04 | |
| D2LatomLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 19.36 | 19.21 | |
| D2LatomLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 18.71 | 18.48 | |
| D2LrawLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 18.67 | 18.57 | |
| D2LrawLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 15.88 | 15.67 |