Search papers, labs, and topics across Lattice.
This study introduces VERA-RL, a novel reinforcement learning framework designed for proactive scientific error verification in academic papers, addressing the limitations of existing models that lack the ability to autonomously identify and justify errors without predefined issues. By constructing the VERA-13K dataset, which encompasses a wide range of scientific error categories, the authors enable models to engage in a Reason-Verify-Scan progression that enhances their reasoning capabilities. Training the Qwen3-VL-8B model with VERA-RL yields significant improvements in verifiable reasoning, achieving performance levels comparable to leading multimodal large language models like Gemini 3 Pro and Qwen3-VL-235B-A22B on the Scan task.
Reinforcement learning can transform how models autonomously verify scientific accuracy, achieving near-human levels of reasoning in error detection.
Multimodal large language models (MLLMs) are increasingly capable scientific assistants, yet they remain far from fully autonomous research. This transition requires models to actively inspect academic papers, build global evidence views, and make traceable judgments without prespecified issues or evidence. However, existing work provides limited task paradigms or training studies for such issue- and evidence-absent verification. We study this challenge through scientific error detection, where models must determine whether errors exist and justify them with evidence-based reasoning. To fill this gap, we present VERA-RL, a reinforcement-learning formulation for scientific error detection over academic papers. Following a Reason--Verify--Scan progression, we construct VERA-13K, a 12,900-sample dataset organized into 4,300 matched chains, covering 6 scientific-error categories across the research workflow and broad natural-science domains. We further introduce fine-grained rewards for reasoning completeness, evidence alignment, and error precision. Training Qwen3-VL-8B with VERA-RL substantially improves verifiable reasoning, approaching flagship MLLMs such as Gemini 3 Pro and Qwen3-VL-235B-A22B on Scan.
