Under a strict identity-level definition, roughly one in twenty 2025 NeurIPS and USENIX Security papers carries at least two likely hallucinated academic citations that survived peer review.
Mapping the increasing use of LLM s in scientific papers
15 Pith papers cite this work, alongside 39 external citations. Polarity classification is still indexing.
15
Pith papers citing it
39
external citations · external index
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
Test-time adaptation with semi-supervised learning leverages inference-time homogeneity to maintain AI text detection performance under adversarial humanization, new LLMs, and temporal drift.
LMs systematically inflate expressed certainty during rewriting, affecting up to 75% of outputs with a 1.5-2x bias toward increasing rather than decreasing certainty, and the effect compounds over iterations.
Introduces a triangulation-based metric to quantify lexical shifts attributable to preference tuning without requiring manual curation of examples.
Luminol-AIDetect detects machine-generated text zero-shot by extracting perplexity-based features from an input and its shuffled version, using density estimation to exploit greater dispersion in MGT perplexity under shuffling.
A think-aloud study reveals that AI tools in early research misrepresent uncertainty, obscure provenance, and create fragile trust, leading researchers to develop compensatory strategies to preserve scholarly judgment.
Introduces CounterBench benchmark and CoIn iterative reasoning method showing LLMs perform near random on formal counterfactual tasks but improve substantially with guided backtracking.
Generative AI use in science can be governed through structured documentation and provenance capture by framing AI interactions as inspectable Research Objects rather than debating authorship.
LLM-associated terms increased in prevalence in full-text scientific articles science-wide from 2021 to 2024, with substantial differences between fields.
A minimal ODE model of AI adoption in writing and reviewing predicts a short-term knowledge peak followed by 40% long-term decline unless review acceleration exceeds writing acceleration.
NLI accuracy on research papers declined steadily over time, with Chinese and French showing unexpected resistance while Japanese and Korean declined more sharply in the post-LLM era.
Presents PAT, an agentic AI review tool using inference scaling that claims 34% better math error recall on SPOT benchmark and successful pilots at STOC and ICML conferences.
An LLM semantic-matching framework for journal recommendation reports 40.23% Top-3 accuracy on 23,609 statistics articles from 49 journals without task-specific training.
AI can generate research artifacts faster than it can verify them, so across all eight lifecycle stages the credible deployment mode is human-governed collaboration rather than full autonomy.
A cross-disciplinary review of 151 studies concludes LLMs accelerate research workflows while introducing recurring technical and ethical risks, including ten it flags as underexplored.
citing papers explorer
-
Phantom References: Hallucinated Citations That Survive Peer Review at Top-Tier Conferences
Under a strict identity-level definition, roughly one in twenty 2025 NeurIPS and USENIX Security papers carries at least two likely hallucinated academic citations that survived peer review.
-
Hitting a Moving Target: Test-Time Adaptation for AI Text Detection under Continual Distribution Shift
Test-time adaptation with semi-supervised learning leverages inference-time homogeneity to maintain AI text detection performance under adversarial humanization, new LLMs, and temporal drift.
-
From `May' to `Is': Certainty Distortion in Language Model Rewriting
LMs systematically inflate expressed certainty during rewriting, affecting up to 75% of outputs with a 1.5-2x bias toward increasing rather than decreasing certainty, and the effect compounds over iterations.
-
Isolating LLM Lexical Bias: A Curation-Free Triangulated Metric for Preference-Stage Learning
Introduces a triangulation-based metric to quantify lexical shifts attributable to preference tuning without requiring manual curation of examples.
-
Luminol-AIDetect: Fast Zero-shot Machine-Generated Text Detection based on Perplexity under Text Shuffling
Luminol-AIDetect detects machine-generated text zero-shot by extracting perplexity-based features from an input and its shuffled version, using density estimation to exploit greater dispersion in MGT perplexity under shuffling.
-
How Researchers Navigate Accountability, Transparency, and Trust When Using AI Tools in Early-Stage Research: A Think-Aloud Study
A think-aloud study reveals that AI tools in early research misrepresent uncertainty, obscure provenance, and create fragile trust, leading researchers to develop compensatory strategies to preserve scholarly judgment.
-
CounterBench: Evaluating and Improving Counterfactual Reasoning in Large Language Models
Introduces CounterBench benchmark and CoIn iterative reasoning method showing LLMs perform near random on formal counterfactual tasks but improve substantially with guided backtracking.
-
Inspectable AI for Science: A Research Object Approach to Generative AI Governance
Generative AI use in science can be governed through structured documentation and provenance capture by framing AI interactions as inspectable Research Objects rather than debating authorship.
-
Have LLM-associated terms increased in article full texts in all fields?
LLM-associated terms increased in prevalence in full-text scientific articles science-wide from 2021 to 2024, with substantial differences between fields.
-
Publish and Perish: How AI-Accelerated Writing Without Proportional Verification Investment Degrades Scientific Knowledge
A minimal ODE model of AI adoption in writing and reviewing predicts a short-term knowledge peak followed by 40% long-term decline unless review acceleration exceeds writing acceleration.
-
Can We Still Hear the Accent? Investigating the Resilience of Native Language Signals in the LLM Era
NLI accuracy on research papers declined steadily over time, with Chinese and French showing unexpected resistance while Japanese and Korean declined more sharply in the post-LLM era.
-
Towards Automating Scientific Review with Google's Paper Assistant Tool
Presents PAT, an agentic AI review tool using inference scaling that claims 34% better math error recall on SPOT benchmark and successful pilots at STOC and ICML conferences.
-
An LLM-Powered Semantic Alignment Framework for Journal Recommendation
An LLM semantic-matching framework for journal recommendation reports 40.23% Top-3 accuracy on 23,609 statistics articles from 49 journals without task-specific training.
-
AI for Auto-Research: Roadmap & User Guide
AI can generate research artifacts faster than it can verify them, so across all eight lifecycle stages the credible deployment mode is human-governed collaboration rather than full autonomy.
-
From Text to Discovery: How Large Language Models Are Reshaping Research Across Scientific and Humanistic Disciplines
A cross-disciplinary review of 151 studies concludes LLMs accelerate research workflows while introducing recurring technical and ethical risks, including ten it flags as underexplored.