HeRo recursively partitions the LLM vocabulary into a hierarchy, embedding multi-bit payloads across layers so that verifiers with different keys recover only their authorized portion while preserving the original sampling distribution.
hub
Release Strategies and the Social Impacts of Language Models
39 Pith papers cite this work. Polarity classification is still indexing.
39
Pith papers citing it
abstract
Large language models have a range of beneficial uses: they can assist in prose, poetry, and programming; analyze dataset biases; and more. However, their flexibility and generative capabilities also raise misuse concerns. This report discusses OpenAI's work related to the release of its GPT-2 language model. It discusses staged release, which allows time between model releases to conduct risk and benefit analyses as model sizes increased. It also discusses ongoing partnership-based research and provides recommendations for better coordination and responsible publication in AI.
hub tools
citation-role summary
background 3
baseline 1
citation-polarity summary
representative citing papers
Unsupervised style representations learned via paraphrase inversion enable competitive few-shot and zero-shot AI-text detection with better generalization to unseen LLMs than supervised baselines.
Steering vectors from frozen LM layers enable a lightweight classifier to detect machine-generated text robustly across domains, source models, and editing attacks.
OpAI-Bench provides a new benchmark for evaluating AI-text detectors on progressively human-AI co-edited documents at multiple granularities, revealing non-monotonic detection patterns.
DEPO formulates detector-evasive paraphrasing as a constrained MDP and solves it via Lagrangian primal-dual RL with GRPO-style updates to achieve evasion while satisfying a semantic-preservation constraint.
A trace-based benchmark of 30 security tasks finds that less-restricted LLM derivatives outperform stock safety-aligned models on some agent tasks for Gemma but not Qwen or Llama, with similar patterns on non-security controls.
ExaGPT uses span-level similarity retrieval from human and LLM datastores to detect machine-generated text while supplying the matching spans as human-interpretable evidence, achieving up to 37-point accuracy gains over prior interpretable detectors at 1% FPR.
LLMs trained on simple specification gaming generalize to zero-shot reward tampering including rewriting their own reward function.
Multitask fine-tuning of an encoder-decoder model on prompted datasets produces zero-shot generalization that often beats models up to 16 times larger on standard benchmarks.
RAG models set new state-of-the-art results on open-domain QA by retrieving Wikipedia passages and conditioning a generative model on them, while also producing more factual text than parametric baselines.
SenFlow integrates graph-based inter-sentence propagation with linear-chain CRF decoding for sentence-level AI-generated text detection and reports SOTA results on the new MOSAIC benchmark of 16k hybrid documents.
DetectZoo is a unified toolkit providing reference implementations of 61 detectors, native loaders for 22 benchmark datasets, and a standardized evaluation pipeline for AI-generated content detection across text, audio, and image modalities.
An image-semantic guided method enhances MLLMs for detecting AI-generated modern Chinese poetry by combining poem text with visual representations of content, achieving 85.65% Macro-F1 with Gemini and outperforming text baselines and RoBERTa.
Steer-to-Detect learns a steering vector injected into LLM hidden states to boost class separability and applies hypothesis testing with finite-sample Type I/II error guarantees for generated-text detection.
MELD is a multi-task AI-text detector using auxiliary heads, uncertainty-weighted losses, EMA distillation, and pairwise ranking that reaches 99.9% TPR at 1% FPR on a new held-out benchmark while remaining competitive on the RAID leaderboard.
A weighted change point detection algorithm locates human- and LLM-written segments in mixed text and attains minimax-optimal localization under heterogeneous sentence scores.
BREW uses block voting and window-shifting verification to reach TPR 0.965 and FPR 0.02 under 10% synonym substitution, addressing high false-positive issues in prior multi-bit LLM watermarking.
DSIPA is a zero-shot black-box detector that uses sentiment distribution consistency and preservation metrics to identify LLM text, reporting up to 49.89% F1 gains over baselines across domains and models.
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.
IRM derives implicit reward signals from off-the-shelf LLMs to detect generated text zero-shot and reports better results than prior zero-shot and supervised detectors on the DetectRL benchmark.
A proposed pipeline shows LLMs introduce detectable race and gender biases when summarizing life narratives, creating potential for representational harm in research.
Open-source AI model development shows far less collaboration and openness than classic open-source software; users innovate by adapting models rather than improving them.
Deeper transformer layers produce intermediate representations that are harder to lossy-compress, and the paper links this to growing covariance and Rademacher complexity.
A human-centered design workshop with journalism practitioners yields an evaluation cookbook and design requirements for contextualized, value-aligned generative AI benchmarks.
citing papers explorer
-
MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text
MELD is a multi-task AI-text detector using auxiliary heads, uncertainty-weighted losses, EMA distillation, and pairwise ranking that reaches 99.9% TPR at 1% FPR on a new held-out benchmark while remaining competitive on the RAID leaderboard.
-
From OSS to Open Source AI: an Exploratory Study of Collaborative Development Paradigm Divergence
Open-source AI model development shows far less collaboration and openness than classic open-source software; users innovate by adapting models rather than improving them.
-
An Overview of Catastrophic AI Risks
The paper categorizes sources of catastrophic AI risks into malicious use, AI race, organizational risks, and rogue AIs, providing illustrative stories and mitigation suggestions for each.