Research map / Large language models / AI and society
Policies Permitting LLM Use for Polishing Peer Reviews Are Currently Not Enforceable
Rounak Saha, Gurusha Juneja, Dayita Chaudhuri, Naveeja Sajeevan, Nihar B Shah, Danish Pruthi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add paper-manuscript context to likelihood computation (ablation) | Fast-DetectGPT · Hard subset AI-BP | TPR (%): 72.1 -> 75.3 | Improved |
| add paper-manuscript context to likelihood computation (ablation) | Fast-DetectGPT · Hard subset H-AI | FPR (%): 4.6 -> 9.5 | Hurt |
| add paper-manuscript context (ablation) | LogLikelihood · Hard subset AI-BP | TPR (%): 46.0 -> 59.8 | Improved |
| add paper-manuscript context (ablation) | LogLikelihood · Hard subset H-AI | FPR (%): 0.0 -> 4.2 | Hurt |
| add paper-manuscript context (ablation) | Binoculars · Hard subset AI-BP | TPR (%): 21.6 -> 20.5 | No clear effect |
| humanize AI-BP reviews with Undetectable AI | Pangram · Easy subset AI-BP | TPR (%): 100 -> 44.6 | Hurt |
| humanize AI-BP reviews | GPTZero · Easy subset AI-BP | TPR (%): 96.7 -> 86.7 | Hurt |
| humanize AI-BP reviews | FastDetect · Easy subset AI-BP | TPR (%): 100 -> 0.0 | Hurt |
| humanize H-AI reviews | GPTZero · Easy subset H-AI | FPR (%): 3.0 -> 65.6 | Hurt |
| humanize H-AI reviews | Pangram · Easy subset H-AI | FPR (%): 3.0 -> 3.63 | Hurt |
| train supervised RoBERTa-base instead of using off-the-shelf detectors (in-distribution) | RoBERTa-base · Hard subset H-AI | FPR (%): 4.3 -> 1.4 | Improved |
| hold out Llama-3.1-70B from supervised training (ablation) | RoBERTa-base · AI-EP | TPR (%): 100.0 -> 87.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Fast-DetectGPT — Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature (ICLR 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews (ICML 2024 · similarity 0.91)
- Position: The Artificial Intelligence and Machine Learning Community Should Adopt a More Transparent and Regulated Peer Review Process (ICML 2025 · similarity 0.88)
- From Replication to Redesign: Exploring Pairwise Comparisons for LLM-Based Peer Review (NeurIPS 2025 · similarity 0.88)
- Position: The AI Conference Peer Review Crisis Demands Author Feedback and Reviewer Rewards (ICML 2025 · similarity 0.87)
- CycleResearcher: Improving Automated Research via Automated Review (ICLR 2025 · similarity 0.87)
- Agent Reviewers: Domain-specific Multimodal Agents with Shared Memory for Paper Review (ICML 2025 · similarity 0.86)
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.
