ToxReason: A Benchmark for Mechanistic Chemical Toxicity Reasoning via Adverse Outcome Pathway

• URL: http://arxiv.org/abs/2604.06264v1
• Date: Tue, 07 Apr 2026 02:32:38 GMT
• Title: ToxReason: A Benchmark for Mechanistic Chemical Toxicity Reasoning via Adverse Outcome Pathway
• Abstract summary: We introduce ToxReason, a benchmark grounded in the Adverse Outcome Pathway (AOP)<n>We evaluate toxicity prediction performance and reasoning quality across diverse large language models (LLMs)<n>We show that reasoning-aware training improves mechanistic reasoning and, consequently, toxicity prediction performance.
• Score: 17.149765900378885
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in large language models (LLMs) have enabled molecular reasoning for property prediction. However, toxicity arises from complex biological mechanisms beyond chemical structure, necessitating mechanistic reasoning for reliable prediction. Despite its importance, current benchmarks fail to systematically evaluate this capability. LLMs can generate fluent but biologically unfaithful explanations, making it difficult to assess whether predicted toxicities are grounded invalid mechanisms. To bridge this gap, we introduce ToxReason, a benchmark grounded in the Adverse Outcome Pathway (AOP) that evaluates organ-level toxicity reasoning across multiple organs. ToxReason integrates experimental drug-target interaction evidence with toxicity labels, requiring models to infer both toxic outcomes and their underlying mechanisms from Molecular Initiating Event (MIE) to Adverse Outcome (AO). Using ToxReason, we evaluate toxicity prediction performance and reasoning quality across diverse LLMs. We find that strong predictive performance does not necessarily imply reliable reasoning. Furthermore, we show that reasoning-aware training improves mechanistic reasoning and, consequently, toxicity prediction performance. Together, these results underscore the necessity of integrating reasoning into both evaluation and training for trustworthy toxicity modeling.

Related papers

• Perception Without Engagement: Dissecting the Causal Discovery Deficit in LMMs [68.58207076756237]
  We introduce ProCauEval, a perturbation-based evaluation protocol that shifts from outcome assessment to mechanism diagnosis.<n>We find that models faithfully perceive video content yet systematically underexploit it during causal reasoning.<n>We propose Anti-Distillation Policy Optimization (ADPO), a reinforcement learning framework built on negative teacher alignment.
  arXiv  Detail & Related papers  (2026-05-10T08:48:58Z)
• LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoning [107.60117957760794]
  We introduce LatentChem, a latent reasoning interface that decouples chemical derivation from textual generation.<n>We show that LatentChem achieves a 59.88% non-tie win rate over strong CoT-based baselines on ChemCoTBench.<n>Our results provide empirical evidence that chemical reasoning is more naturally and effectively realized as continuous latent dynamics.
  arXiv  Detail & Related papers  (2026-02-06T01:28:27Z)
• Rethinking Toxicity Evaluation in Large Language Models: A Multi-Label Perspective [104.09817371557476]
  Large language models (LLMs) have achieved impressive results across a range of natural language processing tasks.<n>Their potential to generate harmful content has raised serious safety concerns.<n>We introduce three novel multi-label benchmarks for toxicity detection.
  arXiv  Detail & Related papers  (2025-10-16T06:50:33Z)
• oMeBench: Towards Robust Benchmarking of LLMs in Organic Mechanism Elucidation and Reasoning [44.36582860924775]
  We introduce oMeBench, the first large-scale, expert-curated benchmark for organic mechanism reasoning in organic chemistry.<n>We also propose oMeS, a dynamic evaluation framework that combines step-level logic and chemical similarity.
  arXiv  Detail & Related papers  (2025-10-09T03:13:31Z)
• <think> So let's replace this phrase with insult... </think> Lessons learned from generation of toxic texts with LLMs [60.169913160819]
  This paper explores the possibility of using synthetic toxic data as an alternative to human-generated data for training models for detoxification.<n>Experiments show that models fine-tuned on synthetic data consistently perform worse than those trained on human data.<n>The root cause is identified as a critical lexical diversity gap: LLMs generate toxic content using a small, repetitive vocabulary of insults that fails to capture the nuances and variety of human toxicity.
  arXiv  Detail & Related papers  (2025-09-10T07:48:24Z)
• CoTox: Chain-of-Thought-Based Molecular Toxicity Reasoning and Prediction [20.143495021407166]
  Large language models (LLMs) offer a promising alternative through step-by-step reasoning and integration of textual data.<n>We propose CoTox, a novel framework that integrates LLM with chain-of-thought (CoT) reasoning for multi-toxicity prediction.<n>Using GPT-4o, we show that CoTox outperforms both traditional machine learning and deep learning model.
  arXiv  Detail & Related papers  (2025-08-05T07:04:44Z)
• Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification? [19.700175505235876]
  ToxiMol is the first benchmark task for general-purpose Multimodal Large Language Models (MLLMs) focused on molecular toxicity repair.<n>We construct a standardized dataset covering 11 primary tasks and 560 representative toxic molecules spanning diverse mechanisms and granularities.
  arXiv  Detail & Related papers  (2025-06-12T17:25:53Z)
• Aligned Probing: Relating Toxic Behavior and Model Internals [78.20380492883022]
  We introduce aligned probing, a novel interpretability framework that aligns the behavior of language models (LMs)<n>Using this framework, we examine over 20 OLMo, Llama, and Mistral models, bridging behavioral and internal perspectives for toxicity for the first time.<n>Our results show that LMs strongly encode information about the toxicity level of inputs and subsequent outputs, particularly in lower layers.
  arXiv  Detail & Related papers  (2025-03-17T17:23:50Z)
• Can LLMs Recognize Toxicity? A Structured Investigation Framework and Toxicity Metric [16.423707276483178]
  We introduce a robust metric grounded on Large Language Models (LLMs) to flexibly measure toxicity according to the given definition. Our results demonstrate outstanding performance in measuring toxicity within verified factors, improving on conventional metrics by 12 points in the F1 score.
  arXiv  Detail & Related papers  (2024-02-10T07:55:27Z)
• Toxicity Detection in Drug Candidates using Simplified Molecular-Input Line-Entry System [0.0]
  The need for analysis of toxicity in new drug candidates and the requirement of doing it fast have asked the consideration of scientists. This paper goes for the study of simplified Molecular Input Line-Entry System (SMILES) as a parameter to develop Long short term memory (LSTM) based models.
  arXiv  Detail & Related papers  (2021-01-21T07:02:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.