Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2608.22972v1 Announce Type: new Abstract: Surgical action triplet recognition constitutes a critical task in context-aware robot-assisted surgery, facilitating automatic surgical action percept...
AI University (AI‑U) is a flexible framework that uses a fine‑tuned large language model (LLM) combined with retrieval‑augmented generation (RAG) and a reasoning synthesis model to produce style‑aligned responses from lecture videos, notes, and textbooks. In a graduate‑level finite‑element‑method (FEM) course, the authors created a pipeline to generate course‑grounded training data, fine‑tuned an open‑source LLM with Low‑Rank Adaptation (LoRA), and applied RAG‑based synthesis. Evaluation through cosine similarity, LLM‑based assessment, expert review, and user studies showed that the expert model outperformed the base model in alignment with course materials, with 86 % of test cases scoring higher and human users preferring the expert model roughly twice as often. whyItMatters":"The study demonstrates a practical method for building course‑specific learning assistants that improve alignment with instructional content, offering a template that can be extended across STEM fields."
arXiv:2608.22950v1 Announce Type: new Abstract: Floating waste in inland waterways threatens aquatic ecosystems and requires timely monitoring under cluttered, multi-object conditions. Existing aquat...
The paper introduces the Multi-Context Fusion Transformer (MFT), a model that predicts pedestrian crossing intentions in urban settings by integrating four types of contextual information—pedestrian behavior, environment, localization, and vehicle motion—through a progressive fusion strategy. MFT uses intra-context attention for reciprocal interactions within each context, cross-context attention to combine these contexts into a global representation, and guided attention mechanisms to refine both context tokens and the global token. Experiments on JAADbeh, JAADall, and PIE datasets show MFT outperforms existing methods with accuracies of 73%, 93%, and 90% respectively, and ablation studies confirm the importance of each network component and input context.
The paper proposes an information‑flow perspective on explainability, arguing that exposing reasons for observed effects is a positive flow of information that must be specified and verified. It introduces an epistemic temporal logic with counterfactual causes to formalize the requirement that agents gain knowledge about why an effect occurred, and presents an algorithm for checking finite‑state models against these specifications. A prototype implementation is evaluated on benchmarks, demonstrating the ability to distinguish explainable from unexplainable systems and to incorporate privacy constraints.
arXiv:2608.21766v1 Announce Type: cross Abstract: Both capability and safety benchmarks rest upon the assumption that the behavior of language models undergoing a test is informative about their beha...
arXiv:2608.22444v1 Announce Type: new Abstract: The unit of AI safety evaluation is still the individual model, yet language-model agents are increasingly deployed in interacting populations that rea...
The paper investigates how large language model agents on the open platform Moltbook represent humans, focusing on human-directed stereotypes. Using an annotation framework with four dimensions—morality, friendliness, competence, and autonomy—and a subtype scheme for other attributions, the study finds that competence is the dominant evaluation, while many other attributions describe humans as epistemic, cultural, or embodied subjects. The authors also analyze how these representations appear in narrative contexts and platform-level circulation, noting that community feedback is better explained by exposure, author visibility, and content selection rather than stable insider–outsider dynamics.
The paper introduces TA-SPA, a black‑box jailbreak method for multimodal large language models that generates transferable perturbations in a text‑anchored semantic space. It combines Text‑Anchored Semantic Factorization (TASF) to separate cross‑modal semantic factors from modality‑specific residuals with Semantic‑Preserving Augmentation (SPA) to diversify harmful target anchors while maintaining semantic consistency. Experiments demonstrate strong attack effectiveness and transferability to commercial MLLMs, with competitive performance against representative defenses.
SynerMedGen is a unified framework that aligns medical multimodal understanding with generation tasks through task alignment. It introduces three generation‑aligned understanding tasks and a two‑stage training strategy that transfers representations learned during understanding to medical image synthesis. The model achieves strong zero‑shot performance on 22 synthesis tasks and outperforms state‑of‑the‑art specialized and unified models when combined with generation training, supported by a new 1M‑sample SynerMed dataset.
The paper investigates how medical large language models (LLMs) may exhibit narrative anchoring bias when presented with the same clinical case in different patient voices. Using the NarrativeShield SDoH MedQA dataset, the authors evaluate three Qwen2.5 instruction‑tuned LLMs (1.5B, 3B, 7B) on 300 clinical cases, reporting metrics such as persona‑level accuracy, counterfactual consistency, correct consistency, and narrative sensitivity error. The 7B model achieves the highest accuracy (56.33 %) and correct consistency (40.33 %), yet narrative sensitivity errors remain substantial (31.67 %).
The article discusses the evolution of AI across three generations—from explicit logic to neural networks to large language models (LLMs)—and how LLMs introduce new systemic risks. It applies a forensic‑psychology profiling method to identify ten key features of LLMs, such as hallucinations, bias, and cognitive atrophy, revealing an entity that confabulates, amplifies user biases, and erodes human competence. The report concludes with a four‑pillar framework for AI resilience, emphasizing cognitive sovereignty, measurable control, partial autonomy, and openness to safeguard society.
DIAG is a Diagnostic Iterative Alignment and Generation framework designed to improve data efficiency in aligning large language models for mathematical reasoning. It adaptively reshapes the practice distribution by first diagnosing valid preference-pair yield to calibrate exploration and exploitation, then generating targeted practice from the model’s failure traces. The approach is theoretically framed as a teacher‑mediated approximation to KL‑regularized reweighting, and experiments show that DIAG increases preference-pair yield and reasoning performance under the same training budget.
arXiv:2608.23248v1 Announce Type: cross Abstract: Traditional clinical prediction models rely on task-specific pipelines and curated, structured data, which scale poorly and underutilize unstructured...
The paper introduces Align‑LoRA, a unified LoRA framework for multi‑task learning that replaces complex, isolated adapter designs with a single‑adapter model enhanced by a higher rank and an explicit alignment loss. It demonstrates that a router‑free, multi‑head model with high inter‑head redundancy can outperform more elaborate baselines, and that a unified LoRA can achieve competitive performance while enabling weight merging and zero inference latency. Extensive experiments and theoretical analysis confirm that Align‑LoRA surpasses prevailing approaches, offering a simpler, production‑friendly paradigm for parameter‑efficient fine‑tuning of large language models.
arXiv:2608.22417v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used to support text analysis in qualitative research, yet evidence on their performance in inductive con...
arXiv:2608.22852v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used in investment decision-making, yet prior work shows that they exhibit systematic, model-specific inv...
TEE-X is a TEE‑aware acceleration framework designed to run large vision models, such as Vision Transformers, entirely within Trusted Execution Environments. It introduces a sensitivity‑aware modularization technique and vectorization to overcome memory constraints and latency challenges on edge devices. The framework is validated on OP‑TEE for Arm TrustZone and optimized for the NVIDIA Jetson AGX Xavier, achieving GPU‑level inference latency with minimal accuracy‑latency trade‑offs.
arXiv:2605.11533v4 Announce Type: replace Abstract: Routine clinical check-up reports combine laboratory measurements, physiological assessments, imaging findings and visually structured information,...
arXiv:2508.13533v2 Announce Type: replace Abstract: Within a model family, a smaller variant is often deployed as a drop-in replacement for a larger one when their performance is similar. However, pe...