Spencer Kim

Aaditya Singh, Adam Fry, Adam Perelman et al. · berkeley, mila

This is the system card published alongside the OpenAI GPT-5 launch, August 2025. GPT-5 is a unified system with a smart and fast model that answers most questions, a deeper reasoning model for harder problems, and a real-time router that quickly decides which model to use based on conversation type, complexity, tool needs, and explicit intent (for example, if you say 'think hard about this' in the prompt). The router is continuously trained on real signals, including when users switch models, preference rates for responses, and measured correctness, improving over time. Once usage limits are reached, a mini version of each model handles remaining queries. This system card focuses primarily on gpt-5-thinking and gpt-5-main, while evaluations for other models are available in the appendix. The GPT-5 system not only outperforms previous models on benchmarks and answers questions more quickly, but -- more importantly -- is more useful for real-world queries. We've made significant advances in reducing hallucinations, improving instruction following, and minimizing sycophancy, and have leveled up GPT-5's performance in three of ChatGPT's most common uses: writing, coding, and health. All of the GPT-5 models additionally feature safe-completions, our latest approach to safety training to prevent disallowed content. Similarly to ChatGPT agent, we have decided to treat gpt-5-thinking as High capability in the Biological and Chemical domain under our Preparedness Framework, activating the associated safeguards. While we do not have definitive evidence that this model could meaningfully help a novice to create severe biological harm -- our defined threshold for High capability -- we have chosen to take a precautionary approach.

Mrinal Agarwal, Saad Rana, Theo Sundoro et al.

Deception is a fundamental challenge for multi-agent reasoning: effective systems must strategically conceal information while detecting misleading behavior in others. Yet most evaluations reduce deception to static classification, ignoring the interactive, adversarial, and longitudinal nature of real deceptive dynamics. Large language models (LLMs) can deceive convincingly but remain weak at detecting deception in peers. We present WOLF, a multi-agent social deduction benchmark based on Werewolf that enables separable measurement of deception production and detection. WOLF embeds role-grounded agents (Villager, Werewolf, Seer, Doctor) in a programmable LangGraph state machine with strict night-day cycles, debate turns, and majority voting. Every statement is a distinct analysis unit, with self-assessed honesty from speakers and peer-rated deceptiveness from others. Deception is categorized via a standardized taxonomy (omission, distortion, fabrication, misdirection), while suspicion scores are longitudinally smoothed to capture both immediate judgments and evolving trust dynamics. Structured logs preserve prompts, outputs, and state transitions for full reproducibility. Across 7,320 statements and 100 runs, Werewolves produce deceptive statements in 31% of turns, while peer detection achieves 71-73% precision with ~52% overall accuracy. Precision is higher for identifying Werewolves, though false positives occur against Villagers. Suspicion toward Werewolves rises from ~52% to over 60% across rounds, while suspicion toward Villagers and the Doctor stabilizes near 44-46%. This divergence shows that extended interaction improves recall against liars without compounding errors against truthful roles. WOLF moves deception evaluation beyond static datasets, offering a dynamic, controlled testbed for measuring deceptive and detective capacity in adversarial multi-agent interaction.

Leo Lu, Jonathan Zhang, Sean Chua et al.

Chain-of-Thought (CoT) prompting has significantly advanced the reasoning capabilities of large language models (LLMs). While prior work focuses on improving model performance through internal reasoning strategies, little is known about the interchangeability of reasoning across different models. In this work, we explore whether a partially completed reasoning chain from one model can be reliably continued by another model, either within the same model family or across families. We achieve this by assessing the sufficiency of intermediate reasoning traces as transferable scaffolds for logical coherence and final answer accuracy. We interpret this interchangeability as a means of examining inference-time trustworthiness, probing whether reasoning remains both coherent and reliable under model substitution. Using token-level log-probability thresholds to truncate reasoning at early, mid, and late stages from our baseline models, Gemma-3-4B-IT and LLaMA-3.1-70B-Instruct, we conduct continuation experiments with Gemma-3-1B-IT and LLaMA-3.1-8B-Instruct to test intra-family and cross-family behaviors. Our evaluation pipeline leverages truncation thresholds with a Process Reward Model (PRM), providing a reproducible framework for assessing reasoning stability via model interchange. Evaluations with a PRM reveal that hybrid reasoning chains often preserve, and in some cases even improve, final accuracy and logical structure. Our findings point towards interchangeability as an emerging behavioral property of reasoning models, offering insights into new paradigms for reliable modular reasoning in collaborative AI systems.

Aayush Aluru, Myra Malik, Samarth Patankar et al.

Multi-agent systems (MAS) often achieve higher reasoning accuracy than single models, but their reliance on repeated debates across agents makes them computationally expensive. We introduce SMAGDi, a distillation framework that transfers the debate dynamics of a five-agent Llama-based MAS into a compact Socratic decomposer-solver student. SMAGDi represents debate traces as directed interaction graphs, where nodes encode intermediate reasoning steps with correctness labels and edges capture continuity and cross-agent influence. The student is trained with a composite objective combining language modeling, graph-based supervision, contrastive reasoning, and embedding alignment to preserve both fluency and structured reasoning. On StrategyQA and MMLU, SMAGDi compresses a 40B multi-agent system into a 6B student while retaining 88% of its accuracy, substantially outperforming prior distillation methods such as MAGDi, standard KD, and fine-tuned baselines. These results highlight that explicitly modeling interaction graphs and Socratic decomposition enable small models to inherit the accuracy benefits of multi-agent debate while remaining efficient enough for real-world deployment.

Hao-Chih Lee, Zelong Liu, Hamza Ahmed et al.

General-purpose vision-language models (VLMs) have emerged as promising tools in radiology, offering zero-shot capabilities that mitigate the need for large labeled datasets. However, in high-stakes domains like diagnostic radiology, these models often lack the discriminative precision required for reliable clinical use. This challenge is compounded by the scarcity and heterogeneity of publicly available volumetric CT datasets, which vary widely in annotation formats and granularity. To address these limitations, we introduce Uniferum, a volumetric VLM that unifies diverse supervision signals, encoded in classification labels and segmentation masks, into a single training framework. By harmonizing three public 3D CT datasets with distinct annotations, Uniferum achieves state-of-the-art performance, improving AUROC on the CT-RATE benchmark by 7% compared to CLIP-based and conventional multi-label convolutional models. The model demonstrates robust out-of-distribution generalization, with observed evidence of unexpected zero-shot performance on the RAD-CHEST and INSPECT datasets. Our results highlight the effectiveness of integrating heterogeneous annotations and body segmentation to enhance model performance, setting a new direction for clinically reliable, data-efficient VLMs in 3D medical imaging.