Research map / Label-efficient and robust learning / Recognition and long-tail
Making a (Counterfactual) Difference One Rationale at a Time
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use counterfactual rather than factual data augmentation | MMI rationale models · TripAdvisor - Location | Rat. Prec.: 31.36 ± 10.38 -> 39.76 ± 10.48 | Helped |
| Use counterfactual augmentation instead of unaugmented MMI training | MMI rationale models · Correlated RateBeer - Appearance | Rat. Prec.: 56.28 ± 17.22 -> 67.79 ± 10.63 | Helped |
| Use counterfactual augmentation instead of unaugmented MMI training | MMI rationale models · Correlated RateBeer - Smell | Rat. Prec.: 45.39 ± 16.45 -> 61.84 ± 6.76 | Helped |
| Use counterfactual augmentation instead of unaugmented MMI training | MMI rationale models · Correlated RateBeer - Palate | Rat. Prec.: 34.98 ± 8.79 -> 41.24 ± 1.72 | Helped |
| Use counterfactual augmentation instead of unaugmented MMI training | MMI rationale models · Decorrelated RateBeer - Appearance | Rat. Prec.: 86.92 ± 7.35 -> 82.82 ± 8.60 | Hurt |
| Use counterfactual instead of factual data augmentation | MMI rationale models · Decorrelated RateBeer - Smell | Rat. Prec.: 83.62 ± 3.52 -> 86.66 ± 2.92 | Helped |
| Use counterfactual instead of factual data augmentation | MMI rationale models · Decorrelated RateBeer - Palate | Rat. Prec.: 66.03 ± 3.76 -> 67.79 ± 3.20 | Helped |
| Use counterfactual rather than factual data augmentation | MMI rationale models · TripAdvisor - Location | Dev. Acc.: 80.61 ± 4.38 -> 78.11 ± 6.77 | Hurt |
Closest papers by meaning
- Iterative Counterfactual Data Augmentation (AAAI 2025 · similarity 0.89)
- Gumbel Counterfactual Generation From Language Models (ICLR 2025 · similarity 0.86)
- A General Search-Based Framework for Generating Textual Counterfactual Explanations (AAAI 2024 · similarity 0.86)
- Counterfactual reasoning: an analysis of in-context emergence (NeurIPS 2025 · similarity 0.84)
- Faithful Explanations of Black-box NLP Models Using LLM-generated Counterfactuals (ICLR 2024 · similarity 0.84)
- NOMATTERXAI: Generating “No Matter What” Alterfactual Examples for Explaining Black-Box Text Classification Models (AAAI 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.