mlsanityData to 14 Oct 2026

Token-Level Likelihood-Array Regression for Membership Inference and AI-Generated Text Detection

arXiv 2608.22179 ↗First posted 23 Aug 2026stat.ML cs.LG stat.MEcited by 0

On the map: LLM safety and security → Membership inference attacks

Abstract (arXiv)

Membership inference asks whether a text was used to train a language model, whereas AI-generated text detection asks whether it was generated by a language model rather than written by a human. Existing likelihood-based methods typically compress token-level probabilities into a few prespecified scores, most often using only probabilities conditioned on the full preceding context. We propose likelihood-array regression (LAR), which evaluates each target token under nested left-context windows and organizes the resulting likelihood-derived features into a structured array. After aligning arrays across texts of different lengths, LAR learns how detection information varies with context scale, token position, and likelihood features. LAR-1 aggregates learned contributions from individual aligned cells, while LAR-2 adds second-order features formed from pairs of evaluations of the same target token across context lengths. For within-path quadratic model, we establish matching minimax lower and upper bounds, characterize errors from finite-dimensional approximation and random squared projections, and derive conditions under which an oracle spectral sieve attains the minimax rate. Across multiple scoring language models, LAR substantially improves membership inference and AI-generated text detection over likelihood-based baselines. The analyses further show that shorter-context likelihoods contain information beyond conventional full-context probabilities, while second-order features provide additional gains for membership inference.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 6 of its 34 references that are in the map. First posted on arXiv 23 Aug 2026.

Builds on

6 of 34 references.

References · 6

  1. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  2. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,411
  3. 2019Release Strategies and the Social Impacts of Language Modelscited by 513
  4. 2023How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation, and Detectioncited by 481
  5. 2025AdaDetectGPT: Adaptive Detection of LLM-Generated Text with Statistical Guaranteescited by 21
  6. 2026Detecting LLM-Generated Text with Performance Guaranteescited by 10

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026OLMo-Detect: A Multi-Stage, Confounder-Controlled Benchmark for Membership Inference on Large Language Models
  2. 2023Did the Neurons Read your Book? Document-level Membership Inference for Large Language Models
  3. 2024ReCaLL: Membership Inference via Relative Conditional Log-Likelihoods
  4. 2024Detecting Training Data of Large Language Models via Expectation Maximization
  5. 2025Tag&Tab: Pretraining Data Detection in Large Language Models Using Keyword-Based Membership Inference Attack
  6. 2024Scaling Up Membership Inference: When and How Attacks Succeed on Large Language Models
  7. 2025Cross-Tokenizer Likelihood Scoring Algorithms for Language Model Distillation
  8. 2026Detecting Data Contamination in Large Language Models