AIarXiv
Ranked by formulaVerdict: NotableMedEvo: Medical Reasoning Evolution through Knowledge Navigation
Medical multimodal reasoning models are typically trained on teacher-generated rationales, discarding the valid reasoning in students' own failed trajectories, while general-domain correction methods offer little…
Key numbers
- 65.57% to 69.12%
- 47.40% to 54.50% on the
- 67.91% to 73.04% and from
Caveats
- Preprint; not yet peer reviewed.
VerdictWorth a reader's time today.
Abstract
Medical multimodal reasoning models are typically trained on teacher-generated rationales, discarding the valid reasoning in students' own failed trajectories, while general-domain correction methods offer little guidance for errors in image reading, anatomical context, or differential diagnosis. We propose MedEvo, which uses medical knowledge, namely a tag taxonomy and error-type diagnosis, to navigate both trajectory repair and data selection. For each failed rollout, a gold-blind auditor locates the first consequential error and provides a minimal, answer-free hint; the same student retains the valid prefix and regenerates the erroneous suffix. After each round, MedEvo identifies capability gaps across medical tags on a held-out feedback set and prioritizes related questions for new rollouts and repairs. With supervised fine-tuning (SFT) alone, MedEvo raises Qwen3.5-9B mean accuracy across seven benchmarks from 65.57% to 69.12% (47.40% to 54.50% on the multimodal track of MedXpertQA (MedXpertQA-MM)), surpassing teacher-distillation SFT (66.93%). It also raises Qwen3.5-35B and Qwen3.5-397B mean accuracy from 67.91% to 73.04% and from 71.83% to 74.95%, respectively. Our analyses show that failed trajectories contain reasoning worth preserving, that repair success does not imply learning value, and that independent feedback is key to sustained improvement.
The editor's rubric
| Dimension | Level | Weight | What that level means |
|---|---|---|---|
| Leverage | ████░ 4 | 24% | A general-purpose tool used across several fields (Adam, ResNet, LoRA, next-generation sequencing). |
| Magnitude | ███░░ 3 | 16% | Large gain: roughly 2x, or a clear new state of the art on a hard, unsaturated problem. |
| Evidence | ███░░ 3 | 20% | Solid: multiple benchmarks or cohorts, ablations, fair baselines, released code or data. |
| Novelty | ██░░░ 2 | 20% | A new combination of known ideas. |
| Trajectory | ██░░░ 2 | 10% | Some room to improve with obvious engineering. |
| Stakes | ██░░░ 2 | 10% | Benefits a professional community (practitioners, clinicians, engineers). |
Editor’s rationale
Heuristic triage from title and abstract text only, not a reading of the paper. Cues found: method (we propose); breadth (many tasks); gains (outperforms); verification (multiple benchmarks, held-out test, independent replication).
How the score was computed
- Merit
- 5.7 / 10
- Adjusted merit
- 5.7 / 10
- Attention
- 0%
- Freshness
- 0%