Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2608. 19621v1 Announce Type: new Abstract: Large language models (LLMs) offer a scalable approach to social simulation, but their credibility depends on how agents are constructed.
arXiv:2608. 19836v1 Announce Type: cross Abstract: Probabilistic shielding is a technique for safe reinforcement learning (RL).
arXiv:2608. 19526v1 Announce Type: cross Abstract: Stock market analysts and investors face a daily challenge: too much financial news, too little time.
arXiv:2605. 24960v2 Announce Type: replace-cross Abstract: Chain-of-Thought (CoT) faithfulness, i.
arXiv:2605. 22547v3 Announce Type: replace-cross Abstract: Medical image diagnosis has achieved significant progress with deep learning, yet existing methods often rely on isolated visual evidence and lack the ability to effectively leverage similar cases and external knowledge.
arXiv:2608. 19748v1 Announce Type: cross Abstract: Inference-time selection methods, such as Best-of-N, improve generation by sampling a pool of candidates and selecting the top-ranked completion according to a reward model.
arXiv:2608. 19964v1 Announce Type: new Abstract: Autonomous driving systems must operate under partial observability, where safety-critical objects may be occluded or visible only to neighboring connected vehicles.
arXiv:2608. 19492v1 Announce Type: new Abstract: World models increasingly treat compact multimodal representations as interfaces between perception and physical interaction, yet existing probes do not establish whether different sensors carry the same executable meaning or whether that meaning survives a new action composition.
arXiv:2608. 19220v1 Announce Type: cross Abstract: Rising immigration has intensified intergroup tensions in many countries.
arXiv:2608. 19425v1 Announce Type: cross Abstract: Reliable performance evaluation is a central bottleneck for deploying robot-learning policies in real-world conditions.
arXiv:2608. 19338v1 Announce Type: cross Abstract: Mechanistic interpretability seeks quantities that models do not expose directly: represented states, component effects, interactions, and responses to interventions.
arXiv:2608. 20011v1 Announce Type: new Abstract: Preference optimization is a standard alignment method for generative models, yet extending it to continuous-time dynamics remains non-trivial.
arXiv:2608. 19861v1 Announce Type: new Abstract: Customer-service LLM agents must follow organizational policy when acting on a user's behalf.
arXiv:2608. 19208v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) are frequently exposed to auxiliary textual context, the impact of which on visually grounded tasks remains underexplored.
arXiv:2608. 19737v1 Announce Type: cross Abstract: Large vision-language models (LVLMs) have achieved remarkable progress in video understanding and reasoning.
arXiv:2608. 20338v1 Announce Type: new Abstract: Large Language Models (LLMs) increasingly require selective removal of harmful or sensitive knowledge, called unlearning, yet existing methods and benchmarks fail to evaluate this capability completely.
arXiv:2608. 20122v1 Announce Type: new Abstract: Large multimodal models (LMMs) have demonstrated strong OCR recognition capabilities, yet remain vulnerable to adversarial visual text that is readable to humans but challenging for models to localize and recognize.
arXiv:2608. 19729v1 Announce Type: new Abstract: Vision-language-model-based embodied agents can complete instructed tasks but often violate safety constraints in the process, a problem recently framed as interactive safety.
arXiv:2608. 19727v1 Announce Type: cross Abstract: Watermarking is a central tool for provenance in generative models, yet its application to multivariate time series remains hindered by reliability failures under post-editing attacks.
arXiv:2608. 19670v1 Announce Type: new Abstract: Large language models (LLMs) compression reduces deployment costs, but standard aggregate metrics like perplexity and accuracy often mask underlying behavioral shifts.