AI reviews for all 22,977 AAAI-26 papers were preferred by authors and PC members over human reviews on accuracy and suggestions and outperformed baselines at spotting weaknesses.
super hub Mixed citations
OpenAI GPT-5 System Card
Mixed citation behavior. Most common role is background (51%).
516
Pith papers citing it
Background
51% of classified citations
abstract
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.
hub tools
citation-role summary
background 44
baseline 23
method 7
dataset 3
other 3
citation-polarity summary
claims ledger
- abstract 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 ar
co-cited works
representative citing papers
TW-LegalBench evaluates 13 LLMs on over 30,000 Taiwanese legal tasks from exams and judgments, showing top models pass lawyer thresholds but struggle with exact statute citations.
A causal audit with image interventions shows text-only models reach within 5.7 accuracy points of top multimodal VLMs on chest radiography, with some large multimodal models statistically indistinguishable from small text-only baselines.
MetaSyn benchmark shows LLM pipelines recover at most 52.7% of ground-truth included studies due to screening failures on PI/ECO eligibility, despite 90.9% retrieval recall at K=200.
EHRNote-ChatQA is the first benchmark for evidence-grounded multi-turn clinical QA over longitudinal discharge summaries, containing 16,072 medical-expert-verified pairs across eight categories and revealing LLM weaknesses in evidence grounding and multi-turn consistency.
Across 30 LLMs and 205 TLA+ tasks, syntactic correctness reaches at most 26.6% and semantic correctness 8.6%, with all successes limited to progressive prompting and no advantage from larger models.
RobotValues is a benchmark of 10K value-conflict scenarios that reveals VLMs default to safety and accommodation while failing to follow instructions to prioritize other values 80% of the time.
Introduces the TVR active viewpoint-matching task and TVRBench indoor simulation benchmark, where foundation models start at low single-digit success rates but reach 51.4% after visual-action SFT and multi-turn GRPO post-training.
AMNESIA is a benchmark suite of 70,560 medical QA pairs that evaluates unlearning methods and shows that patient-level unlearning erodes disease-shared knowledge.
FlowCompile performs compile-time design space exploration on structured LLM workflows to produce reusable high-quality configuration sets that outperform routing baselines with up to 6.4x speedup.
M³Att poisons medical multimodal RAG by pairing covert textual misinformation with query-agnostic visual perturbations that increase retrieval of the bad content, causing LLMs to generate clinically plausible but incorrect responses.
Soohak is a 439-problem mathematician-curated benchmark where frontier LLMs reach at most 30.4% on research math challenges and no model exceeds 50% on refusal for ill-posed problems.
TraceAV-Bench is the first benchmark for multi-hop trajectory reasoning over long audio-visual videos, showing top models reach only 51-68% accuracy with substantial room for improvement.
EVE enables verifiable self-evolution of MLLMs by using a Challenger-Solver architecture to generate dynamic executable visual transformations that produce VQA problems with absolute execution-verified ground truth.
VLMs hallucinate by prioritizing contradictory on-screen text over visual content, addressed via the VisualTextTrap benchmark with 6,057 human-validated samples and the VTHM-MoE dual-encoder framework using dimension-specific experts and adaptive routing.
OccuBench is a new benchmark for AI agents on real-world occupational tasks via LLM-driven simulators, showing no model dominates all industries, implicit faults are hardest, and larger models with more reasoning perform better.
MMRareBench provides 1,756 QA pairs and 7,958 images from PMC rare-disease cases to evaluate 23 MLLMs, revealing low treatment-planning scores and medical models underperforming general models on multi-image tasks due to capacity dilution.
Introduces APRS task and PanoSeeker agent using VLM plus EgoSphere memory for active 360° search and segmentation, outperforming baselines on a new benchmark.
DisciplineGen-1M is a million-scale multidisciplinary dataset for text-to-image generation and editing, paired with a discipline-informed model that improves results on discipline-specific benchmarks.
AnyGroundBench is a domain-adaptation benchmark for spatio-temporal video grounding across animal, industry, sports, surgery, and public security domains that finds 15 state-of-the-art VLMs fail in zero-shot and ICL settings.
LongEgoRefer is a new benchmark of 1,498 referring expressions in 45-minute average egocentric videos that exposes the failure of existing Video REC models on sparse long-form spatio-temporal grounding.
OpenSafeIntent benchmark shows models fail to calibrate safety across intent shifts in matched dual-use prompts, indicating current evaluations are insufficient.
LongVQUBench introduces a hierarchical benchmark with local, cross-event, and global quality understanding tasks plus needle distortion QA to measure LVLMs' long-term video quality reasoning.
An asynchronous architecture decouples incremental voxel-based mapping from VLM-based semantic enrichment to produce queryable open-vocabulary 3D scene graphs that match or exceed prior methods on segmentation and grounding benchmarks.
citing papers explorer
-
When Text Hijacks Vision: Benchmarking and Mitigating Text Overlay-Induced Hallucination in Vision Language Models
VLMs hallucinate by prioritizing contradictory on-screen text over visual content, addressed via the VisualTextTrap benchmark with 6,057 human-validated samples and the VTHM-MoE dual-encoder framework using dimension-specific experts and adaptive routing.
-
On the Cost and Benefit of Chain of Thought: A Learning-Theoretic Perspective
Chain of Thought risk decomposes into oracle-trajectory benefit and trajectory-mismatch cost, with stability determining bounded, linear, or exponential error growth.
-
From Web to Pixels: Bringing Agentic Search into Visual Perception
WebEye benchmark and Pixel-Searcher agent enable visual perception tasks by using web search to resolve object identities before precise localization or answering.
-
Breaking $\textit{Winner-Takes-All}$: Cooperative Policy Optimization Improves Diverse LLM Reasoning
GCPO uses team-level credit assignment via determinant volume over reward-weighted semantic embeddings to promote non-redundant correct reasoning paths, improving both accuracy and diversity in LLM training.
-
MulTaBench: Benchmarking Multimodal Tabular Learning with Text and Image
MulTaBench is a new collection of 40 image-tabular and text-tabular datasets designed to test target-aware representation tuning in multimodal tabular models.
-
StereoTales: A Multilingual Framework for Open-Ended Stereotype Discovery in LLMs
StereoTales shows that all tested LLMs emit harmful stereotypes in open-ended stories, with associations adapting to prompt language and targeting locally salient groups rather than transferring uniformly across languages.
-
Mitigating Many-shot Jailbreak Attacks with One Single Demonstration
A single safety demonstration appended at inference time mitigates many-shot jailbreak attacks by counteracting implicit malicious fine-tuning on harmful examples.
-
Do Joint Audio-Video Generation Models Understand Physics?
AV-Phys Bench shows that current joint audio-video models lack robust physical commonsense, with major drops on transitions and deliberate anti-physics prompts.
-
Self-Driving Datasets: From 20 Million Papers to Nuanced Biomedical Knowledge at Scale
Starling, a multi-agent LLM system, extracts ~6.3 million nuanced structured records from PubMed across six tasks with reported error rates of 0.6-7.7%, lower than several curated databases.
-
ClassEval-Pro: A Cross-Domain Benchmark for Class-Level Code Generation
ClassEval-Pro benchmark shows frontier LLMs achieve at most 45.6% Pass@1 on class-level code tasks, with logic errors (56%) and dependency errors (38%) as dominant failure modes.
-
Benchmarking and Improving GUI Agents in High-Dynamic Environments
DynamicUI improves GUI agent performance in high-dynamic environments by processing interaction videos with frame clustering, action-conditioned refinement, and reflection, outperforming prior approaches on the new DynamicGUIBench spanning ten applications.
-
A systematic evaluation of vision-language models for observational astronomical reasoning tasks
Vision-language models underperform specialized astronomical methods on real observational data, with accuracy improving when physical explanations are provided in prompts and when raw numerical measurements replace rendered plots.
-
Using large language models for embodied planning introduces systematic safety risks
LLM planners for robots often produce dangerous plans even when planning succeeds, with safety awareness staying flat as model scale improves planning ability.
-
Psychological Steering of Large Language Models
Mean-difference residual stream injections outperform personality prompting for OCEAN trait steering in most LLMs, with hybrids performing best and showing approximate linearity but non-human trait covariances.
-
Mosaic: Cross-Modal Clustering for Efficient Video Understanding
Mosaic uses cross-modal clusters as the unit for KVCache organization in VLMs to achieve up to 1.38x speedup in streaming long-video understanding.
-
QAPruner: Quantization-Aware Vision Token Pruning for Multimodal Large Language Models
QAPruner introduces a hybrid sensitivity metric that combines group-wise quantization error simulation and outlier intensity with semantic scores to prune visual tokens, yielding 2.24% higher accuracy than naive baselines at 12.5% token retention on LLaVA models while surpassing dense low-bit models
-
Reasoning4Sciences: Bridging Reasoning Language Models to All Scientific Branches
A survey of RLM use in 28 disciplines reveals uneven adoption and introduces a maturity assessment framework showing larger gaps when limited to public resources.
-
Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making
Frontier LLMs exhibit bias from stigmatizing language in clinical vignettes across four conditions, skewing decisions toward less aggressive management, with limited mitigation from Chain-of-Thought or self-debiasing prompts.
-
The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space
Reformulating 53 visual reasoning tasks in polar coordinates causes frontier MLLMs to drop from 70-83% to 31-39% accuracy while preserving logical equivalence, revealing a Cartesian shortcut in current benchmarks.
-
When to Re-Commit: Temporal Abstraction Discovery for Long-Horizon Vision-Language Reasoning
Learns state-conditioned commitment depth in a 7B vision-language policy that jointly predicts actions and replan intervals, outperforming fixed-depth baselines and larger models on Sliding Puzzle and Sokoban while providing a theoretical dominance result.
-
MT-JailBench: A Modular Benchmark for Understanding Multi-Turn Jailbreak Attacks
MT-JailBench is a modular benchmark that standardizes evaluation of multi-turn jailbreaks to identify key success drivers and enable stronger combined attacks.
-
Object Hallucination-Free Reinforcement Unlearning for Vision-Language Models
HFRU is a two-stage reinforcement unlearning method operating on the vision encoder with GRPO optimization and an abstraction reward that achieves over 98% forgetting and retention on object and face tasks with negligible hallucination.
-
Chart-FR1: Visual Focus-Driven Fine-Grained Reasoning on Dense Charts
Chart-FR1 uses Focus-CoT for linking reasoning to visual cues and Focus-GRPO reinforcement learning with efficiency rewards to outperform prior MLLMs on dense chart reasoning tasks.
-
DenseStep2M: A Scalable, Training-Free Pipeline for Dense Instructional Video Annotation
A scalable training-free pipeline using video segmentation, filtering, and off-the-shelf multimodal models creates DenseStep2M, a dataset of 100K videos and 2M detailed instructional steps that improves dense captioning, step grounding, and cross-modal retrieval.
-
Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation
Lightning OPD is an offline on-policy distillation method that matches standard OPD performance at 4x efficiency by enforcing teacher consistency between SFT and distillation phases.
-
When LLMs Lag Behind: Knowledge Conflicts from Evolving APIs in Code Generation
LLMs produce executable code only 42.55% of the time under API evolution without full documentation, improving to 66.36% with structured docs and by 11% more with reasoning strategies, yet outdated patterns persist.
-
Let Geometry GUIDE: Layer-wise Unrolling of Geometric Priors in Multimodal LLMs
GUIDE unrolls multi-granularity geometric priors layer-wise into early MLLM layers with gating to improve spatial reasoning and perception.
-
An Independent Safety Evaluation of Kimi K2.5
Kimi K2.5 matches closed models on dual-use tasks but refuses fewer CBRNE requests and shows some sabotage and self-replication tendencies.
-
Memory in the LLM Era: Modular Architectures and Strategies in a Unified Framework
A unified framework for LLM agent memory is benchmarked, with a new hybrid method outperforming state-of-the-art on standard tasks.
-
From Data to Theory: Autonomous Large Language Model Agents for Materials Science
An LLM agent autonomously selects, codes, and validates materials equations from data, recovering known laws reliably but requiring checks for new or specialized cases.
-
What Limits Vision-and-Language Navigation ?
StereoNav reaches new benchmark highs on R2R-CE and RxR-CE and improves real-robot reliability by supplying persistent target-location priors and stereo-derived geometry that stay stable under lighting changes and blur.
-
CoT-Guard: Small Models for Strong Monitoring
CoT-Guard is a 4B model using SFT and RL that achieves 75% G-mean^2 on hidden objective detection under prompt and code manipulation attacks, outperforming several larger models.
-
SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture
SenseNova-U1 presents native unified multimodal models that match top understanding VLMs while delivering strong performance in image generation, infographics, and interleaved tasks via the NEO-unify architecture.
-
Fill the GAP: A Granular Alignment Paradigm for Visual Reasoning in Multimodal Large Language Models
GAP aligns visual latent reasoning in MLLMs via PCA-mapped decoder outputs, auxiliary visual supervision, and selective capacity-guided training, yielding top supervised performance on a 7B model with evidence that latents carry task-relevant signal.
-
A Composite Activation Function for Learning Stable Binary Representations
HTAF is a sigmoid-tanh composite that approximates the Heaviside function to allow stable gradient training of binary activation networks, yielding ICBMs with stable discretization and competitive performance on image tasks.
-
Evaluating the False Trust Engendered by LLM Explanations
LLM reasoning traces and post-hoc explanations increase false trust in incorrect predictions, whereas contrastive dual explanations enhance users' ability to distinguish correct from incorrect AI outputs.
-
Ace-Skill: Bootstrapping Multimodal Agents with Prioritized and Clustered Evolution
Ace-Skill boosts multimodal agent self-evolution via prioritized rollouts with lazy-decay tracking and semantic knowledge clustering, yielding up to 35% relative gains on tool-use benchmarks and zero-shot transfer to smaller models.
-
How You Begin is How You Reason: Driving Exploration in RLVR via Prefix-Tuned Priors
IMAX trains soft prefixes with an InfoMax reward to drive diverse exploration in RLVR, yielding up to 11.60% gains in Pass@4 over standard RLVR across model scales.
-
Do Agents Need to Plan Step-by-Step? Rethinking Planning Horizon in Data-Centric Tool Calling
Full-horizon planning with on-demand replanning achieves accuracy parity with single-step planning in tool-calling agents for knowledge base and multi-hop question answering while consuming 2-3 times fewer tokens.
-
Minimizing Modality Gap from the Input Side: Your Speech LLM Can Be a Prosody-Aware Text LLM
TextPro-SLM reduces the speech-text modality gap by feeding an LLM backbone with synchronized text tokens and prosody embeddings from WhisperPro, achieving lowest gap scores at 3B/7B scales with roughly 1,000 hours of audio.
-
UniEP: Unified Expert-Parallel MoE MegaKernel for LLM Training
UniEP fuses MoE communication and computation into unified MegaKernels with deterministic token ordering, delivering 1.03x-1.38x speedups over prior work while preserving training accuracy.
-
CoEnv: Driving Embodied Multi-Agent Collaboration via Compositional Environment
CoEnv introduces a compositional environment that integrates real and simulated spaces for multi-agent robotic collaboration, using real-to-sim reconstruction, VLM action synthesis, and validated sim-to-real transfer to achieve high success rates on multi-arm manipulation tasks.
-
DAT: Dual-Aware Adaptive Transmission for Efficient Multimodal LLM Inference in Edge-Cloud Systems
DAT combines a small-large model cascade with fine-tuning and bandwidth-aware multi-stream transmission to deliver high-accuracy event recognition and low-latency alerts for video streams in edge-cloud systems.
- Next-Scale Autoregressive Models for Text-to-Motion Generation