# NovGauge: A Fine-Grained Benchmark for Diagnosing LLMs' Capability in Paper Novelty Assessment

arXiv: https://arxiv.org/abs/2609.11234 · HTML: https://mlsanity.com/paper/2609.11234/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 52 references that are in the map.
- First posted on arXiv 10 Sep 2026.

## Builds on (most cited first, 24 of 24)

- [SPECTER: Document-level Representation Learning using Citation-informed Transformers](https://mlsanity.com/paper/2004.07180/) (2020)
- [Can large language models provide useful feedback on research papers? A large-scale empirical analysis](https://mlsanity.com/paper/2310.01783/) (2023)
- [Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews](https://mlsanity.com/paper/2403.07183/) (2024)
- [CycleResearcher: Improving Automated Research via Automated Review](https://mlsanity.com/paper/2411.00816/) (2024)
- [SciRepEval: A Multi-Format Benchmark for Scientific Document Representations](https://mlsanity.com/paper/2211.13308/) (2022)
- [MARG: Multi-Agent Review Generation for Scientific Papers](https://mlsanity.com/paper/2401.04259/) (2024)
- [DeepReview: Improving LLM-based Paper Review with Human-like Deep Thinking Process](https://mlsanity.com/paper/2503.08569/) (2025)
- [Evaluating and Enhancing Large Language Models for Novelty Assessment in Scholarly Publications](https://mlsanity.com/paper/2409.16605/) (2024)
- [Mind the Blind Spots: A Focus-Level Evaluation Framework for LLM Reviews](https://mlsanity.com/paper/2502.17086/) (2025)
- [Literature-Grounded Novelty Assessment of Scientific Ideas](https://mlsanity.com/paper/2506.22026/) (2025)
- [HindSight: Evaluating LLM-Generated Research Ideas via Future Impact](https://mlsanity.com/paper/2603.15164/) (2026)
- [Stop Automating Peer Review Without Rigorous Evaluation](https://mlsanity.com/paper/2605.03202/) (2026)
- [Generative Adversarial Reviews: When LLMs Become the Critic](https://mlsanity.com/paper/2412.10415/) (2024)
- [ScholarEval: Research Idea Evaluation Grounded in Literature](https://mlsanity.com/paper/2510.16234/) (2025)
- [From Replication to Redesign: Exploring Pairwise Comparisons for LLM-Based Peer Review](https://mlsanity.com/paper/2506.11343/) (2025)
- [Insights from the ICLR Peer Review and Rebuttal Process](https://mlsanity.com/paper/2511.15462/) (2025)
- [Is this Idea Novel? An Automated Benchmark for Judgment of Research Ideas](https://mlsanity.com/paper/2603.10303/) (2026)
- [InnoEval: On Research Idea Evaluation as a Knowledge-Grounded, Multi-Perspective Reasoning Problem](https://mlsanity.com/paper/2602.14367/) (2026)
- [NovBench: Evaluating Large Language Models on Academic Paper Novelty Assessment](https://mlsanity.com/paper/2604.11543/) (2026)
- [Can AI Be a Good Peer Reviewer? A Survey of Peer Review Process, Evaluation, and the Future](https://mlsanity.com/paper/2604.27924/) (2026)
- [Navigating Ideation Space: Decomposed Conceptual Representations for Positioning Scientific Ideas](https://mlsanity.com/paper/2601.08901/) (2026)
- [Detecting AI-Generated Content in Academic Peer Reviews](https://mlsanity.com/paper/2602.00319/) (2026)
- [Beyond Rating: A Comprehensive Evaluation and Benchmark for AI Reviews](https://mlsanity.com/paper/2604.19502/) (2026)
- [From Isolated Scoring to Collaborative Ranking: A Comparison-Native Framework for LLM-Based Paper Evaluation](https://mlsanity.com/paper/2603.17588/) (2026)

## Abstract (arXiv)

Large language models (LLMs) are increasingly used in peer review at major AI conferences, yet novelty remains a persistent weak point. Existing benchmarks assess novelty as a single holistic score, making it difficult to diagnose which dimension a model misjudges or whether its evidence is faithful. We present NovGauge, a human-anchored benchmark for fine-grained novelty assessment diagnosis. The benchmark contains 619 paper pairs and 50 multi-paper sets, drawn from two expert sources: ICLR reviewer overlap claims and survey co-citations. Instances are independently labeled along three dimensions: task, problem, and method, capturing application goals, technical challenges, and solution approaches. We propose a cascading diagnostic pipeline that verifies per-dimension correctness, evidence grounding, and logical support. Evaluation of 18 LLMs shows hallucination rates ranging from 0% to 39% across dimensions, and among non-hallucinated correct-positive judgments, over 70% cite evidence fails to logically support the stated reason. The best-performing model, GPT-5.5, achieves 43-72% Verified F1 across dimensions, while most models retain less than half of their raw F1 after faithfulness verification. These results suggest that current LLMs remain far from reliable scientific novelty assessment, particularly when correctness is conditioned on faithful evidence grounding.