NeurIPS 2026 Papers

LLM-based systems are increasingly deployed in domains where incorrect outputs carry real costs, often necessitating expensive human review and verification. We argue that a central issue is not simply error, but _consequence-insensitivity_: models fail to adjust their behavior according to the cost of being wrong. We study this failure across both explicit utility framings as well as natural-language descriptions of stakes that mirror real-world deployment scenarios. Our evaluation spans agentic coding and mathematical reasoning, covering five frontier model families. Across settings, models systematically under-abstain: they continue to answer or submit patches even when incorrect answers carry significant consequences. The pathology is striking: models continue to submit answers in trivial settings where abstention is strictly dominant, and even when told an incorrect answer will cause nuclear extinction. Further, we find that this behavior is orthogonal to existing benchmarks; increasing model size or capability within a family does not improve sensitivity. Comparing base and instruction-tuned models localizes much of this anti-abstention bias to post-training: base models abstain far more often, though they are not themselves reliably consequence-aware. Finally, we evaluate prior post-hoc interventions alongside in-context learning and fine-tuning, finding that none robustly resolves the failure. Our results suggest that current post-training and evaluation pipelines optimize models to answer, not to act under stakes -- a significant bottleneck for trustworthy autonomy.

While recent vision-language models (VLMs) have achieved significant improvements on static visual-to-code tasks such as generating code for webpages, charts, or SVGs, it is still unclear VLMs are able to recover temporal dynamics when motion is present. To this end, we introduce the animation-to-code task, as well as a benchmark, \texttt{Animation2Code}, for evaluating temporal visual reasoning via reconstructing executable web animation code from videos. \texttt{Animation2Code} consists of 1,069 web animation videos with diverse visuals and motion patterns, augmented with corresponding HTML/CSS/JavaScript implementations. We further pair the dataset with several novel human-aligned metrics, which allow us to disentangle visual similarity from temporal alignment when comparing rendered animations to ground-truth samples. We benchmark current SOTA model performance on these new samples, and show that current models struggle with temporal consistency, even when achieving high appearance similarity, even in fine-tuning and iterative refinement scenarios. Our benchmark dashboard is available at: \url{https://anim2code-dashboard.vercel.app}.

While Large Language Models (LLMs) have significantly advanced radiology report generation (RRG) with strong generative priors, standard autoregressive decoders still suffer from sequential error accumulation when they must infer all clinical structures directly from images. Diffusion Language Models (DLMs) have recently emerged as a non-autoregressive paradigm with iterative masking and denoising mechanisms. Instead of replacing the mature language decoder with a DLM, we reveal that the key value of DLMs for RRG lies in controllable clinical planning, where the model predicts globally consistent anchor points before free-text realization. To harness this complementary strength, we propose ANCHOR-GEN, the first framework that uses a DLM as an explicit diffusion language planner for RRG. Rather than forcing the autoregressive decoder to infer every critical structure from visual tokens alone, ANCHOR-GEN predicts a structured clinical canvas that serves as clinically grounded anchor points. An autoregressive LLM then leverages these anchors to complete report generation with both fluency and clinical faithfulness. By systematically combining non-autoregressive planning and autoregressive decoding, our unified pipeline is trained with structured denoising, hierarchical coarse-to-fine supervision, and planner-decoder binding objectives. Extensive experiments on the public \textsc{MIMIC-CXR} benchmark demonstrate that ANCHOR-GEN improves clinical faithfulness over prior works. Comprehensive analyses and ablations further show that diffusion planning and causal decoding offer complementary strengths for RRG, and that integrating them in one pipeline provides an effective route toward more faithful medical text generation.

Analog circuit design remains highly dependent on expert knowledge due to the complexity of device-level interactions and topology design. Recent transformer-based approaches for device-level topology generation have shown promise, yet they suffer from low electrical validity without human-in-the-loop (HITL) training and severe memorization caused by sequence-based circuit representations. In this work, we propose **AnalogToBi**, a framework for device-level analog circuit topology generation. AnalogToBi introduces *circuit-type conditioning* for categorizing heterogeneous multi-type topology datasets, *device renaming augmentation* to mitigate memorization, *a bipartite graph representation* for improved structural generalization, and *grammar-guided decoding* to enforce structural validity during bipartite graph generation. Experimental results demonstrate that AnalogToBi achieves high validity and novelty without HITL training while effectively avoiding memorization of training topologies.

Generative modeling of longitudinal Electronic Health Records is increasingly important for privacy-preserving research, yet standard autoregressive models tend to underrepresent the co-occurrence structure of tail events (i.e., diseases, symptoms), reducing the fidelity and faithfulness of generated data for rare subpopulations. To this end, we propose ADAPCLA framework, which enables generative models to adaptively fit and generate EHR data through a data distribution-aware training strategy; this is achieved by internalizing data knowledge parameters by simulated annealing training. It also supports training-free adaptation to a diverse clinical population for generation through zero-shot distribution control. Moreover, our theoretical analysis characterizes rare-code logit updates through the label-wise empirical NTK and derives a prior-internalization bound for how annealing speed and NTK conditioning affect retained prior signals. Experiments on real-world data show that ADAPCLA achieves consistent gains in tail plausibility, downstream utility, and zero-shot control; in particular, it improves TailPairSeen over HALO by 114.2\% on MIMIC-III and 65.1\% on MIMIC-IV, outperforms GPT-style generation by 3.5\% F1 for zero-shot cross-population adaptation.

Current LLM-for-GPU benchmarks largely measure CUDA fluency on isolated operators or short kernel-generation tasks, whereas practical GPU performance engineering often starts from trusted sequential CPU programs and requires correct, end-to-end acceleration. To fill this gap, we introduce \textsc{AccelEval}, a benchmark for whole-program CPU-to-CUDA acceleration. AccelEval contains 42 tasks derived from public repositories and established HPC/application suites across six domains, each equipped with deterministic correctness oracles and three input scales. Beyond pass/fail correctness, AccelEval reports end-to-end speedup over CPU baselines and a novel peak-attainment metric which normalizes each model’s speedup against the best correct solution. Using this framework, we evaluate eight frontier LLMs across multiple input scales, GPU architectures and sample rates. At pass@1 on NVIDIA H200 at medium scale, the strongest model passes 38/42 tasks at 71.2$\times$ geometric-mean speedup over the CPU baseline. On the subset of 11 tasks with human-written CUDA references, the best-of-8 LLM oracle reaches 0.75$\times$ of human CUDA performance. To make CUDA-optimization behavior interpretable and reusable, we decompose correct solutions into 43 CUDA acceleration patterns and show that feeding winner-derived pattern guidance back to models improves paired geometric-mean speedup by 1.78$\times$, compared with 1.18$\times$ for a generic-tips control. Together, these results establish AccelEval as a reproducible diagnostic platform for end-to-end LLM-assisted GPU acceleration. An anonymized open-source release is available at \url{https://anonymous.4open.science/r/AccelEval-2397}.

Designing antibodies that recognize a target antigen while avoiding non-cognate antigens remains a central challenge in computational antibody engineering. Existing antigen-conditioned generative models can jointly design antibody sequences and structures, but they are typically trained to match structural data distributions and lack an explicit alignment mechanism for antigen-specific preferences. We present AbSpecAlign, a plug-and-play specificity-reward alignment framework for antibody design. AbSpecAlign introduces a Bidirectional Multimodal Specificity Scorer (BiMSS), which integrates antibody and antigen sequence representations with SE(3)-equivariant structural embeddings to produce a contrastive reward over cognate versus non-cognate antigen pairs. We further introduce a GRPO-based post-training strategy that uses group-relative specificity rewards to align pre-trained antibody generators toward higher-scoring antigen-conditioned designs while maintaining structural regularity. To the best of our knowledge, AbSpecAlign is among the first frameworks to apply GRPO-style group-relative reward alignment to antigen-conditioned antibody sequence--structure co-design with an explicit off-target specificity reward. Experiments on RAbD CDR design benchmarks show that AbSpecAlign improves most sequence-recovery and structural-fidelity metrics when combined with diffusion-based antibody generators. Experiment results suggest that GRPO-based reward alignment with contrastive specificity signals provides an effective post-training mechanism for antigen-conditioned antibody generation, while prospective experimental validation remains necessary.

Typical LLM responses tend to follow a default style, even though users often have distinct preferences regarding tone, verbosity, and formality that they do not explicitly state in their prompts. Evaluating whether personalization methods can adapt to these implicit preferences is challenging, since users typically provide prompts rather than reference responses, style preferences are not factually verifiable, and reference-free LLM judges may conflate personalization with general response quality. To address these challenges, we introduce the Arbitrary Preference Mapping (APM) benchmark, which decouples user attributes (e.g. enthusiastic) from response principles (e.g. persuasive) via a hidden, randomized mapping $\mathbf{C}$ that maps user attributes to preferences about response traits. Because $\mathbf{C}$ carries no semantic content and is resampled across runs, models cannot exploit stereotypical associations and must infer preferences from conversation history. Using this unbiased evaluation methodology, we adapt retrieval-augmented, prompt-optimization, and routing personalization methods and evaluate them on Llama-3.1-8B and Qwen-3.5-27B. Our results show that routing is the most reliable approach, while RAG benefits only from the stronger base LLM, and soft prompt optimization fails to improve significantly over a non-personalized baseline. Our extensive evaluation reveals that in this realistic setting, personalization remains challenging, but our adapted methods show promise.

We study language generation in the limit under a global preference ordering on strings, as introduced by Kleinberg and Wei. We aim for breadth, but impose an additional requirement of timeliness: higher-ranked strings should be generated earlier. A string is then only credited if it is generated before a deadline, where its deadline is defined by a function that maps a string’s rank in the target language to the time by which it must be produced. This is in keeping with a central consideration in machine learning, where inductive bias favors "simpler" or "more plausible" outputs, all else being equal. We show that timely generation is impossible in a strong sense for eventually consistent generators—the protagonists of most prior related work. Under what is perhaps the mildest natural relaxation of consistency, a hallucination rate that vanishes over time, we show that we can circumvent our impossibility result. In particular, we can achieve optimal density with respect to any superlinear deadline function. We also show this is tight by ruling out timely generation with linear deadlines and vanishing hallucination rate.

Code assistants are increasingly utilized in test-driven software development, yet the theoretical mechanisms behind their environment-interaction strategies remain underexplored. We provide a probabilistic framework for two dominant paradigms: code selection after generation using the execution environment, and code generation conditioned on environment feedback. First, we formalize several well-established selection heuristics as environment-aware estimators of code correctness. We theoretically prove that estimators based on fuzzy functional similarity add an inductive bias and strictly dominate estimators based on functional equivalence in terms of signal-to-noise ratio. Second, we frame backprompting as an in-context approximation of Thompson sampling. We derive a novel regret bound for reward functions with unobservable components, theoretically explaining why the effectiveness of backprompting is limited by the ambiguity of the informal task description (an irreducible regret). Using five state-of-the-art open weight models, we corroborate these findings across BigCodeBenchHard, LeetCodeDataset, and QiskitHumanEvalSim. Our formalization also suggests how to improve task descriptions effectively, leading to a new benchmark, QiskitHumanEvalSimX.

The Objective Structured Clinical Examination (OSCE) is widely used to assess clinical communication skills in medical education, requiring trainees to take histories and reason clinically with standardized patients (SPs). Large language model (LLM)-based virtual standardized patients (VSPs) offer a practical alternative to human SPs. These are typically built using persona-injected prompting, which casts a model into a clinical role via a single system prompt. However, existing evaluations of such VSPs often rely on a single backbone LLM, small human rater panels, or LLM-as-a-judge scoring without an utterance-aligned human-expert reference. Consequently, they are limited in their ability to identify the specific shortcomings of individual models. In this study, we evaluated four LLMs (GPT-4o, Claude 3.5 Sonnet, Llama-3.1 70B, and HyperCLOVA X), spanning commercial, open-source, and Korean-developed models, against a human-expert reference across three Korean OSCE cases (hematuria, fever in pregnancy, and chronic cough). In a double-blind, fully crossed design, 17 senior medical-student raters scored every model utterance on three sentence-level and six encounter-level Likert metrics, yielding 7,480 data points. We analyzed the scores using a stratified linear mixed-effects model based on an 11-category history-taking and 3-level linguistic-form taxonomy. At the encounter level, three of the four LLMs fell 0.08 to 0.47 Likert points below the human-expert reference on Comprehensibility, Consistency and Hallucination-freeness. The difference was not statistically significant for Hallucination-freeness in Claude 3.5 Sonnet and GPT-4o, Consistency in GPT-4o, and Engagement in HyperCLOVA X. Llama-3.1 70B fell significantly below the reference on all six metrics. Furthermore, we introduce a web-based OSCE platform that supports virtual practice across diverse clinical cases. This platform continuously collects student–VSP dialogues under research and privacy consent; the resulting dataset is available at https://huggingface.co/datasets/neurips-2026-cpx/neurips-2026-cpx

Dynamic Scene Graph Generation (DSGG) aims to capture objects and their evolving relations in videos. Despite recent progress, the practicality and quality of generated scene graphs remain limited compared to the rapid advances in Multimodal Large Language Models (MLLMs). In this work, we revisit DSGG from two fundamental perspectives: task setup and model design. From the task setup perspective, we identify two key limitations of the current recall-oriented evaluation protocol: (i) a severe precision–recall trade-off, and (ii) uninformative and redundant relation generation. To better assess practical usefulness, we introduce four additional metrics that provide a more comprehensive evaluation of the quality of generated dynamic scene graphs. From the model design perspective, we explore directly using MLLMs for scene graph generation and establish a strong MLLM-based DSGG baseline through three design changes. First, we replace the conventional bottom-up pipeline with a top-down reason-then-locate strategy. Second, we reformulate frame-wise dynamic graphs as Temporal Relation Set (TRS) prediction, improving both efficiency and performance. Third, we introduce Importance-Aware Finetuning (IAF) to encourage more relevant and diverse relation generation. Extensive experiments on Action Genome, VidVRD, and PVSG show that our approach consistently achieves state-of-the-art performance across multiple metrics.

Procedural 3D modeling through code is emerging as a versatile paradigm, offering deterministic, engine-ready, and precisely editable assets that neural 3D generators inherently lack. Authoring such procedural content, however, demands deep expertise in 3D software APIs, parametric design, and code-level geometric reasoning. In this paper, we propose 3DCodeBench, a systematic benchmark for evaluating vision-language model (VLM) agents for procedural 3D generation in 3D modeling software. Specifically, 3DCodeBench evaluates how effectively 11 advanced VLMs can serve as procedural 3D modelers by transferring text and image references into procedural code for software. Recognizing that automated metrics may not fully capture the perceptual quality of 3D shapes, we build 3DCodeArena, a ranking platform based on human preference between generated pairs. From extensive evaluations and results, we observe that: (1) Failures mostly arise from API usage, while successful renders still suffer from disconnected or floating 3D geometric components. (2) Test-time scaling, such as higher thinking budgets and multi-turn refinement, improves performance overall. Our findings highlight a critical need for high-quality procedural coding data to advance commercial VLMs. Furthermore, effective procedural 3D modeling requires a robust execution environment that provides high-fidelity feedback for iterative refinement. To accelerate research in this domain, we release 3DCodeBench—including the curated data, benchmark, evaluation protocol, and qualitative results—as a foundational tool for developing VLM-based 3D modelers.

Generating property-optimized mRNA sequences is central to applications such as vaccine design and protein replacement therapy, but remains challenging due to limited data, complex sequence-function relationships, and the narrow space of biologically viable sequences. Generative methods that drift away from the data manifold can yield sequences that fail to fold, translate poorly, or are otherwise nonfunctional. We present RNAGenScape, a property-guided manifold Langevin dynamics framework for mRNA sequence generation that operates directly on a learned manifold of real data. By performing iterative local optimization constrained to this manifold, RNAGenScape preserves biological viability, accesses reliable guidance, and avoids excursions into nonfunctional regions of the ambient sequence space. The framework integrates three components: (1) an autoencoder jointly trained with a property predictor to learn a property-organized latent manifold, (2) a denoising autoencoder that projects updates back onto the manifold, and (3) a property-guided Langevin dynamics procedure that performs optimization along the manifold. Across three real-world mRNA datasets spanning two orders of magnitude in size, RNAGenScape increases median property gain by up to 148% and success rate by up to 30% while ensuring biological viability of generated sequences, and achieves competitive inference efficiency relative to existing generative approaches.

Vehicle trajectory generation is critical for autonomous driving simulation. However, current generative models face two coupled limitations: (i) imbalanced trajectory distributions bias them toward frequent motions, such as straight driving or stationary states, while weakening their ability to synthesize rare but safety critical motions, such as turning. (ii) Euclidean intermediate spaces provide limited geometric inductive bias for modeling vehicle pose evolution. Together, these limitations lead to mode biased generation and physically less coherent trajectories. To address this issue, we propose SEA, an \underline{\textbf{SE}}(2)-\underline{\textbf{A}}ware vehicle trajectory generation method that realizes the conditional transport path through planar rigid body poses. At the \textbf{data distribution} level, SEA learns vehicle trajectory distributions from traffic data to capture individual motion characteristics. At the \textbf{physical-space} level, motivated by the geometry dependence of {\bf optimal transport} interpolation, SEA realizes intermediate transport states through an $SE(2)$-consistent construction, thereby preserving geometric interpretability and physical coherence during generation. Extensive experiments on large scale real world datasets show that our method outperforms state of the art approaches by generating more realistic, diverse and dynamically consistent trajectories.