A new arXiv preprint examines recurrent graph neural networks, which repeatedly run message passing until it converges. The authors note that existing logical characterizations of these models rely on multi-set aggregation, counting-based graded logics, and halting or acceptance conditions that cannot be verified from the network's parameters.
A revised preprint on arXiv examines how the structure of a lexicon relates to the compositional way sentence meanings are built from word meanings. The author notes that much existing work on semantic universals assumes either fixed signal structures in lexicons or holistic composition that cannot be interpreted. The work frames these questions within evolutionary semantics.
Tree-based models such as random forests tend to assign higher importance scores to continuous predictors than to categorical ones. The authors analyze the source of this bias theoretically and suggest a straightforward fix: injecting a small amount of noise into each categorical predictor. The work targets mixed data settings where such scoring imbalance is common.
A arXiv paper examines where the concept of animacy is encoded inside large language models. The authors trace internal circuits tied to the animate/inanimate distinction, which involves verb-argument constraints and contextual cues beyond simple word-level features. The work is a revised version of a preprint in the cs.CL category.
A preprint proposes a framework for electronic Neuromorphic Design Automation, described as a pipeline connecting computational neuroscience models with established electronic design automation practices. The authors argue for a unified, end-to-end approach rather than treating the two domains separately. The work is a revised cross-listing on arXiv and presents a conceptual design flow rather than a released tool.
A new arXiv preprint presents a mesh-based super-resolution method that uses graph transformers to reconstruct multiscale detonation flow data. The authors argue such data-driven reconstruction is useful for subgrid closure modeling, faster spatiotemporal forecasting, compression, and as an upsampling step in simulations. The work appears as a cross-listed revision in the cs.AI and cs.LG categories.
A new arXiv preprint describes an automated system that turns textbook PDFs into interactive video explanations tailored to individual questions. After a user uploads a PDF and submits a query, the pipeline produces a video answer, with the reported focus on NCERT educational materials. The work targets personalized, question-driven learning content.
A revised arXiv paper introduces WaveHiTS, a wavelet-enhanced hierarchical time series model designed for short-term wind direction forecasting in eastern Inner Mongolia. The approach targets common difficulties in directional data, including circular values, multi-step error accumulation, and complex meteorological interactions. It was cross-listed on arXiv's cs.LG and cs.AI categories.
A new arXiv preprint introduces a method for checking the safety of agentic network operations (NetOps) systems at the level of individual actions rather than coarse task outcomes. The work targets autonomous networks that adjust workloads and respond to incidents, where verification granularity matters for reliability. The authors argue that finer-grained safety signals are needed before such agents can be trusted in production networks.
A new arXiv paper introduces the Perceptual Reality Transformer, a model aimed at helping people convey atypical perceptual experiences while keeping their intended meaning intact. The work argues that recognizable imagery alone is insufficient, since such accounts also carry vividness, duration, uncertainty, and emotional tone. It examines what an illustration needs to retain so those qualities survive translation into a generated image.
A new arXiv paper proposes Semantic-TVM, a virtual memory design that keeps sensitive values protected while still letting agent workflows run on remote language models. The approach targets memory-augmented and tool-using agents, where retrieved memories, tool calls, and intermediate observations can leak private data. It aims to move past one-way masking, which hides values but also blocks the trusted execution they are needed for.
A new arXiv paper examines how AI agents that run long workflows through external tools can leave inconsistent state even when individual tool calls report success. It focuses on conditions such as retries, speculative execution, concurrency, and partial failures. The work frames these mismatches as anomalies at the boundary between the agent and the tools it calls.
A new arXiv preprint examines how enterprise AI agents are typically given static credentials at deployment that mirror the full set of permissions an employee role could hold. The authors argue this approach, inherited from role-based access control, grants agents far more access than any single task requires. They evaluate an alternative architecture that scopes an agent's permissions to the specific task it is performing.
A new arXiv paper introduces Lightning Weave, a method aimed at pushing the accuracy-efficiency frontier of reasoning models. The authors argue that accuracy and inference efficiency often pull toward different reasoning behaviors, making joint improvement difficult. Their approach relies on composing capabilities rather than optimizing the two objectives independently.
A new arXiv paper introduces DiffAdapterVLA, a method that folds continuous trajectory planning directly into the backbone of a pretrained driving vision-language model. The authors argue that while driving VLMs absorb rich visual, route, language and driving context, their training objectives stay disconnected from continuous planning, so prior work tends to bolt planning on separately. The approach targets native continuous trajectory generation rather than a detached planning head.
A new arXiv preprint introduces RSIAgent, a multi-agent system that lets digital agents explore unfamiliar environments and iteratively improve themselves without any additional training. The approach is designed for settings where interfaces, tools, and failure patterns differ from what pretrained models have seen. The work appears under arXiv:2609.15364v1 in both cs.AI and cs.CL.
A new arXiv paper introduces SkillLift, a method for improving the reusable procedural prompts that LLM agents keep as persistent skills, which lets them adapt without retraining model weights. Rather than rewriting skill text directly from execution feedback, the approach derives dense scoring rubrics from limited, costly oracle evaluations to make skill evolution more efficient. The work targets lower evaluation cost during agent skill self-improvement.
A new arXiv paper proposes combining Interacting Multiple Model (IMM) filtering with a state prediction neural network for multiple object tracking in autonomous driving. The approach targets radar sensing, which keeps working in poor weather and provides relative velocity via the Doppler effect. The work aims to improve tracking robustness for obstacle avoidance and route planning.
A new arXiv preprint argues that enterprise AI agents frequently work in demos but fail once they are asked to run continuously in production. The author attributes this to pilots that never ship and to deployed systems that discard feedback instead of learning from it, and proposes an approach called substrate inversion to address the gap. The work is categorized under cs.AI and is cross-listed on arXiv.
A new arXiv paper addresses the challenge of computing controlled invariant sets in real time for safety-critical autonomous control, where conditions shift during operation. Standard synthesis methods scale poorly as state dimension grows, so the authors exploit monotone system structure to derive robust invariant sets more efficiently. The work targets applications needing formal safety certificates computed online rather than offline.
A new arXiv paper proposes conflict-predictive variable horizons for distributed model predictive control used in multi-drone collision avoidance. Fixed horizons force a trade-off between computational cost and early reaction to nearby drones, while the proposed method adjusts the horizon based on predicted conflicts. The approach aims to keep control cheap while still responding in time to approaching neighbors.
A new arXiv paper introduces DA-DLM, a method for diffusion language models that explicitly captures relationships between tokens during generation. Existing diffusion models denoise masked text by predicting several tokens at once under an assumption of conditional independence, which the authors say weakens coherence. The proposed approach aims to restore those inter-token dependencies.
Researchers propose ProtoCAM, a method that combines prototypical networks with mask guidance to classify breast lesions in ultrasound images using only a small number of labeled examples. The approach is designed to be interpretable, addressing the difficulty of building reliable deep learning models for ultrasound analysis, especially in patients with dense breast tissue. The work is published as an arXiv preprint.
A new arXiv preprint introduces BusMA, a communication substrate intended to coordinate multi-agent systems that handle planning, tool use, and evidence synthesis. The authors argue that current designs, which rely on hierarchical manager-worker structures or router-based message passing, have limitations that a bus-style architecture could address. The paper is a preprint and has not yet been peer reviewed.
A new arXiv preprint describes a method for corporate and financial-disclosure analysis that avoids training and cross-model alignment steps. The authors argue that dense text embeddings and large language models struggle with context limits, hallucination risk, compute cost, and inconsistent vector spaces across independently trained models. The work is demonstrated on SEC filings, with a revised version now posted.
This arXiv paper notes that intraoral scan datasets such as Bite2Text arrive already aligned in occlusion, which means key occlusal measurements can be calculated directly rather than inferred by a learned captioning model. The authors argue for deriving these quantities in closed form as a basis for automatically generating orthodontic reports. The work falls under computation and language research and has not been peer reviewed.
A new arXiv preprint proposes "Discovery Foundation Models," framing open-ended discovery as the next stage for AI systems. The authors argue that models have moved from recalling and reasoning over existing knowledge to acting with tools and learning from outcomes, and that the next step is generating genuinely new findings. The paper is a position piece rather than an experimental release.
A new arXiv preprint argues that the two most common query-document scoring methods, TF-IDF and BM25, both correspond exactly to Kullback-Leibler divergences. This gives the widely used retrieval heuristics a probabilistic justification and places them inside a single statistical framework. The author frames the work as filling a long-standing gap in how these ranking functions are derived.
A revised arXiv preprint introduces a formal framework for theory of mind that derives an agent's beliefs from the history of information they encountered, rather than assuming those beliefs are already known. The account models factors such as the order of exposure, the source of information, and its perceived credibility. This extends prior formal treatments that generally take beliefs as given inputs.
A new arXiv preprint treats software vulnerability detection as a binary classification task and proposes a fusion approach guided by Fisher information to combine natural code sequence representations with other modalities. The method aims to adaptively weight each modality's contribution rather than fusing them uniformly. The work targets improved accuracy in flagging code snippets that contain security defects.
A new arXiv paper proposes a method for converting natural-language specifications into SysMLv2, the textual modeling language standardized for model-based systems engineering. The approach refines candidate translations iteratively, using conformance checks to guide corrections. It aims to lower the barrier to producing formal system models that capture requirements, structure, and behavior.
Researchers propose a certified, differentiable complexity estimator that extends algorithmic information dynamics beyond the piecewise-constant Block Decomposition Method. This allows gradient-based control of complexity during training, which they apply to the phenomenon of grokking. The work is published as an arXiv preprint in machine learning.
A new arXiv preprint addresses a weakness in outcome-based reinforcement learning for language-model agents: because the whole trajectory receives a single advantage signal, individual decisions get only coarse credit over long interaction sequences. The authors propose reconciling process supervision with outcome-based credit, drawing on on-policy self-distillation to produce finer-grained guidance. The work is presented as a revised submission and targets long-horizon agent training.
A new arXiv paper introduces Learning to Coach (L2C), a framework that trains a separate LLM acting as a coach to pull actionable guidance out of experience. The motivation is that raw solution trajectories are typically long and noisy, which limits how well language models can learn from them. The approach aims to convert such trajectories into more useful, condensed coaching signals.
A new arXiv paper examines end-to-end supervised graph prediction, where comparing predicted graphs to targets with arbitrary node orderings requires a permutation-invariant loss. The authors analyze three optimal transport-based graph matching relaxations and propose amortization to reduce the cost of these matching steps.
A new arXiv paper proposes a retrieve-localize-generate pipeline for retrieval-augmented generation aimed at answering questions over long-term conversation history. The authors argue that current RAG methods fall short in this setting and frame their approach as addressing those gaps. The work appears as a replacement submission on arXiv's computation and language listing.
A new preprint sets out a framework for describing when a learned or acquired representation broadens what a system can do, and where that broadening stops. Its central claim is that a system can gain new cognitive capabilities from a representation without inheriting the abilities that were used to create it. The work aims to make the scope and limits of such capability gains specifiable.
A new arXiv preprint addresses the difficulty of synthesizing nonverbal vocalizations such as laughs, sighs, and coughs in controllable speech generation. The authors attribute the challenge to the acoustic variety of these sounds and their uneven representation in existing speech corpora, and propose modeling, scaling, and decoding methods to improve them. The work is listed under the cs.AI cross-list announcement.
A new arXiv preprint studies "moral rebellion," the idea that an autonomous agent may choose to disobey assigned tasks when they conflict with moral obligations encountered during execution. The work frames this as a decision-making problem arising from competing duties rather than a simple failure to comply. It appears in the cs.AI category as a new submission.
Researchers introduce LePlanner, an iterative amortized controller designed to plan within the latent spaces of world models built on joint-embedding predictive architectures. The work targets the high computational cost of existing planning methods, which typically rely on either search-based or gradient-based optimization. The paper is posted on arXiv as a cross-listing.