Pith. sign in

REVIEW 29 cited by

ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2203.09509 v4 pith:444SPTCJ submitted 2022-03-17 cs.CL

ToxiGen: A Large-Scale Machine-Generated Dataset for Adversarial and Implicit Hate Speech Detection

classification cs.CL
keywords toxictoxigentextlanguagemachine-generateddatagroupshate
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Toxic language detection systems often falsely flag text that contains minority group mentions as toxic, as those groups are often the targets of online hate. Such over-reliance on spurious correlations also causes systems to struggle with detecting implicitly toxic language. To help mitigate these issues, we create ToxiGen, a new large-scale and machine-generated dataset of 274k toxic and benign statements about 13 minority groups. We develop a demonstration-based prompting framework and an adversarial classifier-in-the-loop decoding method to generate subtly toxic and benign text with a massive pretrained language model. Controlling machine generation in this way allows ToxiGen to cover implicitly toxic text at a larger scale, and about more demographic groups, than previous resources of human-written text. We conduct a human evaluation on a challenging subset of ToxiGen and find that annotators struggle to distinguish machine-generated text from human-written language. We also find that 94.5% of toxic examples are labeled as hate speech by human annotators. Using three publicly-available datasets, we show that finetuning a toxicity classifier on our data improves its performance on human-written data substantially. We also demonstrate that ToxiGen can be used to fight machine-generated toxicity as finetuning improves the classifier significantly on our evaluation subset. Our code and data can be found at https://github.com/microsoft/ToxiGen.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 29 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety

    cs.CL 2026-05 unverdicted novelty 7.0

    Boiling the Frog is a new stateful multi-turn benchmark for agentic safety that reports an aggregate strict attack success rate of 44.4% across nine models, with rates ranging from 20.5% to 92.9% depending on the mode...

  2. Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety

    cs.CL 2026-05 unverdicted novelty 7.0

    Boiling the Frog is a new stateful multi-turn benchmark that finds an aggregate 44.4% strict attack success rate for incremental safety violations across nine AI models, with rates ranging from 20.5% to 92.9%.

  3. Distributionally Robust Multi-Task Reinforcement Learning via Adaptive Task Sampling

    cs.LG 2026-05 unverdicted novelty 7.0

    DRATS derives a minimax objective from a feasibility formulation of MTRL to adaptively sample tasks with the largest return gaps, leading to better worst-task performance on MetaWorld benchmarks.

  4. The Safety-Aware Denoiser for Text Diffusion Models

    cs.LG 2026-04 unverdicted novelty 7.0

    SAD modifies the denoising process in text diffusion models to enforce safety constraints at inference time, reducing unsafe generations while preserving quality and diversity.

  5. MHSafeEval: Role-Aware Interaction-Level Evaluation of Mental Health Safety in Large Language Models

    cs.CL 2026-04 unverdicted novelty 7.0

    MHSafeEval applies a new role-aware taxonomy to discover cumulative mental health harms in LLM counseling trajectories via adversarial multi-turn interactions, revealing failures missed by static benchmarks.

  6. FlowGuard: Towards Lightweight In-Generation Safety Detection for Diffusion Models via Linear Latent Decoding

    cs.CV 2026-04 unverdicted novelty 7.0

    FlowGuard detects unsafe content during diffusion image generation via linear latent decoding and curriculum learning, outperforming prior methods by over 30% F1 while reducing GPU memory by 97% and projection time to...

  7. Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression

    cs.CL 2025-02 unverdicted novelty 7.0

    KV cache compression causes task-dependent degradation in high-density reasoning due to disrupted CoT links; ShotKV mitigates this by preserving few-shot examples as indivisible semantic units through phase separation...

  8. Symmetric Divergence and Normalized Similarity: A Unified Topological Framework for Representation Analysis

    stat.ML 2026-06 unverdicted novelty 6.0

    Introduces SRTD and SRTD-lite to symmetrize topological divergences for neural representations and NTS as a rank-correlation-based metric bounded in [-1,1] for cross-scenario benchmarking.

  9. SafetyRepro: Configuration-Conditional Rank Instability on Alignment Benchmarks

    cs.LG 2026-05 unverdicted novelty 6.0

    Configuration choices alone flip pairwise safety verdicts on every tested alignment benchmark, isolated via a finite-envelope proposition linking disagreement rate to strict ordering reversal.

  10. S2T-RLHF: Hierarchical Credit Assignment for Stable Preference-Based RLHF

    cs.AI 2026-05 conditional novelty 6.0

    S2T-RLHF splits each response-level RLHF reward into sentence shares and then token shares, via bargaining and Dirichlet weighting, yielding steadier training with competitive preference alignment.

  11. Leveraging RAG for Training-Free Alignment of LLMs

    cs.LG 2026-05 unverdicted novelty 6.0

    RAG-Pref is a training-free RAG-based alignment technique that conditions LLMs on contrastive preference samples during inference, yielding over 3.7x average improvement in agentic attack refusals when combined with o...

  12. Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks

    cs.AI 2026-05 unverdicted novelty 6.0

    Toxicity benchmarks for LLMs produce inconsistent results when task type, input domain, or model changes, revealing intrinsic evaluation biases.

  13. f-GRPO and Beyond: Divergence-Based Reinforcement Learning Algorithms for General LLM Alignment

    cs.LG 2026-02 unverdicted novelty 6.0

    f-GRPO and f-HAL estimate f-divergences between reward-aligned and reward-unaligned response distributions and prove expected reward improvement for general LLM alignment.

  14. Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises Safety

    cs.CL 2025-12 unverdicted novelty 6.0

    Distilling safe refusal behavior from OpenAI o1-mini into Llama-3, Gemma-2, and Qwen3 models via response-based LoRA on multilingual jailbreak data increases jailbreak success rates on MultiJail by up to 16.6 points.

  15. Red-Bandit: Test-Time Adaptation for LLM Red-Teaming via Bandit-Guided LoRA Experts

    cs.CL 2025-10 unverdicted novelty 6.0

    Red-Bandit adapts online to LLM failure modes by dynamically selecting among RL-trained LoRA attack-style experts via a bandit policy, reporting SOTA ASR@10 on AdvBench with lower-perplexity prompts.

  16. SmoothLLM: Defending Large Language Models Against Jailbreaking Attacks

    cs.LG 2023-10 accept novelty 6.0

    SmoothLLM mitigates jailbreaking attacks on LLMs by randomly perturbing multiple copies of a prompt at the character level and aggregating the outputs to detect adversarial inputs.

  17. Textbooks Are All You Need II: phi-1.5 technical report

    cs.CL 2023-09 unverdicted novelty 6.0

    phi-1.5 is a 1.3B parameter model trained on synthetic textbook data that matches the reasoning performance of models five times larger on natural language, math, and basic coding tasks.

  18. AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models

    cs.CL 2023-04 accept novelty 6.0

    AGIEval shows GPT-4 exceeding average human scores on SAT Math at 95% and Chinese college entrance English at 92.5%, while revealing weaker results on complex reasoning tasks.

  19. UniSteer: Text-Guided Flow Matching in Activation Space for Versatile LLM Steering

    cs.CL 2026-05 unverdicted novelty 5.0

    UniSteer trains a conditional flow matching model on LLM residual-stream activations to enable text-conditioned steering and classification across multiple behavioral tasks.

  20. Fence: Specialized SLM Guardrails for LLM Applications

    cs.AI 2026-05 conditional novelty 5.0

    Small language models trained on GAN-style adversarial synthetic data outperform prompted large models on use-case-specific guardrails for off-topic and prompt-injection detection.

  21. The Safety-Aware Denoiser for Text Diffusion Models

    cs.LG 2026-04 unverdicted novelty 5.0

    Safety-Aware Denoiser integrates safety guidance into the denoising steps of text diffusion models to reduce unsafe generations while maintaining quality.

  22. DRAFT: Task Decoupled Latent Reasoning for Agent Safety

    cs.LG 2026-02 unverdicted novelty 5.0

    DRAFT decouples agent safety judgment into latent extraction and reasoning stages, raising average benchmark accuracy from 63.27% to 91.18%.

  23. AgentCrypt: Advancing Privacy and (Secure) Computation in AI Agent Collaboration

    cs.CR 2025-12 unverdicted novelty 5.0

    AgentCrypt introduces a deterministic three-tier privacy framework for AI agent collaboration that uses masking and homomorphic encryption to protect data independently of model accuracy.

  24. TrustLLM: Trustworthiness in Large Language Models

    cs.CL 2024-01 unverdicted novelty 5.0

    TrustLLM defines eight trustworthiness principles, creates a six-dimension benchmark, and evaluates 16 LLMs showing proprietary models generally lead but some open-source ones are close while over-calibration can hurt...

  25. Distilling Safe LLM Systems via Soft Prompts for On Device Settings

    cs.LG 2026-06 unverdicted novelty 4.0

    Soft prompt distillation with total variation and KL divergence transfers safety behaviors from guard models to on-device LLMs and outperforms LoRA adapters, steering vectors, and direct optimization in safety-usefuln...

  26. Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions

    cs.CY 2026-02 reject novelty 4.0

    A position paper argues that post-hoc XAI explanations are unfaithful and paradoxical, proposing a shift to expert-based verification and certification of AI systems.

  27. Beyond Explainable AI (XAI): An Overdue Paradigm Shift and Post-XAI Research Directions

    cs.CY 2026-02 unverdicted novelty 4.0

    Current XAI methods for DNNs and LLMs rest on paradoxes and false assumptions that demand a paradigm shift to verification protocols, scientific foundations, context-aware design, and faithful model analysis rather th...

  28. Baichuan 2: Open Large-scale Language Models

    cs.CL 2023-09 unverdicted novelty 4.0

    Baichuan 2 presents 7B and 13B LLMs trained on 2.6T tokens that match or exceed similar open models on MMLU, CMMLU, GSM8K, HumanEval and excel in medicine and law.

  29. AERIC: Anticipatory Hidden-State Monitoring for Implicit Harmful Dialogue

    cs.CL 2026-05 unverdicted novelty 3.0

    AERIC uses a 387-parameter head on LLM hidden states for same-pass anticipatory detection of implicit harm, reporting AUROC gains on DiaSafety and Harmful Advice plus low-latency trigger rates on HarmBench and SocialH...