Short-context Reading Comprehension on DROP (F1, ROUGE-L)
45.18F1 ScoreBase+ctx
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Base+ctxLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 45.18 | 45.18 | |
| Base+ctxLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 42.61 | 42.61 | |
| D2LatomLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 30.66 | 30.66 | |
| D2LckptLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 29.62 | 29.55 | |
| D2LatomLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 29.52 | 29.52 | |
| Doc2AtomLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 28.31 | 28.31 | |
| D2LrawLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 27.12 | 27.12 | |
| Doc2AtomLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 27.05 | 27.05 | |
| D2LrawLLM Backbone=Gemma-2-2B-It, Evaluation Protocol=In-domain2026.06 | 26.98 | 26.98 | |
| D2LckptLLM Backbone=Qwen3-4B-Instruct, Evaluation Protocol=In-domain2026.06 | 10.07 | 10.02 |