AI Research & Papers — September 21, 2026

Findings

[1] 2026-09-21 Improving synthesis prediction of small molecules at scale with RetroChimera

At a glance We report on the recent publication of our retrosynthesis model RetroChimera in the journal Nature (opens in new tab). The paper describes the model’s architecture as well as extensive validation studies, including the model’s ability to recall… As a result, RetroChimera performs strongly across both common and rare reaction classes and produces retrosynthesis predictions that better align with chemists’ judgment (Figure 3). In blind tests, expert chemists preferred disconnections of complex molecules suggested by RetroChimera over those…
[2] 2026-09-21 A Generative Grammar Underlying the Voynich Manuscript, the Pastiche Hypothesis: Evidence from Large Language Models

arXiv:2609.20835v1 Announce Type: new Abstract: Background: The Voynich Manuscript is a fifteenth-century codex written in an unknown script whose content remains undeciphered. Previous studies suggest that its statistical properties resemble those of natural languages, while its illustrations – primarily plants – recall medieval herbals. Methods: We present a multidisciplinary analysis combining probabilistic modeling, phonetic decomposition, rare-event detection, and multimodal image…
[3] 2026-09-21 Offline Multimodal Large Language Models for Decision Support in Air Operations

arXiv:2609.21390v1 Announce Type: cross Abstract: Air operations rely on complex rules, established procedures, and time-critical analysis under limited connectivity and strict security constraints. In such environments, analysts must combine written doctrine with images, often without access to external computing resources. This paper studies offline large language models as decision support tools, deployed in isolated and restricted environments to give analysts…
[4] 2026-09-21 Towards Secure Cloud-Native Computing: Unveiling Kubernetes Misconfigurations with Large Language Models

arXiv:2609.20834v1 Announce Type: new Abstract: In the rapidly evolving landscape of cloud-native computing, Organizations are increasingly adopting infrastructure models that emphasize scalability, flexibility, and efficiency. Kubernetes has become the de facto standard for orchestrating containerized applications in these environments. However, the inherent complexity of cloud-native ecosystems introduces significant challenges, particularly in the form of misconfigurations that can compromise both security…
[5] 2026-09-21 Disentangling Long-Term Memory via Latent Neuro-Symbolic Reasoning

arXiv:2609.18461v2 Announce Type: replace-cross Abstract: Personalized agents are required to reason over long-term history interactions to infer both explicit preferences and implicit behavioral evidence. While early flat retrieval methods score memory fragments independently and neglect the distributed information, current structured memory frameworks rely on query-agnostic static graphs that fail to capture the context-dependent relations. Crucially, raw textual memories are inherently…
[6] 2026-09-21 Transsion's Speaker-Attributed Multilingual ASR System for the MLC-SLM 2026 Challenge

arXiv:2609.20833v1 Announce Type: new Abstract: This paper presents the Transsion Speech Team submission to Task 1 of the MLC-SLM 2026 Challenge, which focuses on speaker-attributed transcription for multilingual conversational speech. We propose a cascaded framework consisting of three components: a speaker diarization module, a long-form multilingual ASR module, and a speaker-transcription fusion module. The diarization module is built upon DiariZen…
[7] 2026-09-21 MENASpeechBank: A Reference Voice Bank with Persona-Conditioned Multi-Turn Conversations for AudioLLMs

arXiv:2602.07036v2 Announce Type: replace-cross Abstract: Audio large language models (AudioLLMs) enable instruction following over speech and general audio, but progress is limited by the scarcity of diverse, conversational, and instruction-aligned speech–text data. This gap is particularly pronounced for persona-grounded and dialectal interactions, where collecting real multi-speaker recordings remains costly and slow. We introduce MENASpeechBank, a reference speech bank comprising ~18K…
[8] 2026-09-21 TALON: A Temporally Aware Longitudinal Framework for Radiology Report Generation

arXiv:2609.20826v1 Announce Type: new Abstract: Current radiology report generation (RRG) models usually produce descriptive reports based on a single examination or only the most recent prior examination, limiting their ability to perform accurate and meaningful longitudinal comparisons and detect subtle interval changes. Although recent approaches have begun to incorporate multiple prior examinations, they usually aggregate a fixed-length history without explicitly…
[9] 2026-09-21 From Discharge Notes to Patient Understanding: Persona-Grounded, Open-Ended Simulation of LLMs as Discharge Educators

arXiv:2609.20827v1 Announce Type: new Abstract: Hospital discharge education is an interactive teaching task: a clinician adapts a discharge plan to a patient's literacy, recall, and personality. Existing LLM evaluations target static or artifact-generation tasks and do not measure patient understanding under open-ended dialogue. We introduce DischargeBench, a persona-grounded simulation in which a candidate LLM educator conducts a multi-turn session with…
[10] 2026-09-21 Beyond WER: Entity and Disfluency Recall in Accented Conversational ASR

arXiv:2609.20828v1 Announce Type: new Abstract: ASR systems optimised for Word Error Rate (WER) often miss named entities and filled pauses in accented conversational English, both critical for language-learning feedback. We present a three-stage pipeline for speakers from India, Indonesia, and Latin America: (1) heuristic SQL filters curating entity-rich training data at 2.8x the entity density of random sampling, (2) regional…
[11] 2026-09-21 Reviser: Revision-Capable Text Generation via Autoregressive Cursor Actions

arXiv:2609.20830v1 Announce Type: new Abstract: Revision-capable generation is appealing because it can insert or revise earlier content, but many non-autoregressive and edit-based approaches obtain this flexibility through repeated sequence-level computation. We propose Reviser, a decoder-only Transformer that generates a response as a sequence of cursor-relative actions on a mutable canvas. At each step, Reviser predicts exactly one action token: INSERT(token),…
[12] 2026-09-21 SAGE: Schema-Guided LLMs for Grant Review

arXiv:2609.20829v1 Announce Type: new Abstract: Grant reviewers must apply detailed criteria to application forms, budgets, and supporting documents while producing assessments that colleagues can inspect. We present SAGE, Schema-Guided Aspect-Based Grant Evaluation, a system that translates a grant rubric into structured checks and links its judgements to evidence from the application package. We evaluate SAGE in two stages on 35…
[13] 2026-09-21 TatBLiMP: A Benchmark of Linguistic Minimal Pairs for Tatar

arXiv:2609.20832v1 Announce Type: new Abstract: We introduce TatBLiMP, the first benchmark of linguistic minimal pairs for Tatar (tt, ISO 639-3 tat), a Qypchaq Turkic language written in Cyrillic. To our knowledge it is the first grammaticality evaluation for Tatar language models of any kind, since even the 101-language MultiBLiMP does not include Tatar. TatBLiMP covers 16 morphosyntactic phenomena in 1248…
[14] 2026-09-21 Recursive Language Models Generalize Out of Domain

arXiv:2609.20831v1 Announce Type: new Abstract: We study when limiting what a language model can see improves learning. We compare standard CoT, the more general learner that reads the full trace, with recursive language models, which restricts itself by solving each subtask in an isolated context. In-distribution, this generality comes for free: CoT can efficiently simulate the recursive rule, so the…
[15] 2026-09-21 The Spoken Wikipedia Presentation Corpus

arXiv:2609.21676v1 Announce Type: cross Abstract: We present the Spoken Wikipedia Presentation Corpus, an extension of the Spoken Wikipedia Corpora featuring LLM-generated slide decks for multimodal ASR. Slides are created from LLM-segmented sections using a hybrid pipeline that combines LLM-based content planning with rule-based design decisions. For each section, an LLM generates a slide title, bullet points, a takeaway message, and…
[16] 2026-09-21 Clinician-Grounded Quality Assurance for AI-Assisted Psychiatric Intake

arXiv:2609.21149v1 Announce Type: cross Abstract: Before patients can use AI-assisted psychiatric intake systems, health systems need practical ways to routinely evaluate these tools against their clinical standards for quality assurance. Because clinicians may use different intake styles, evaluation for this task must (1) support comparison across interviewing approaches, (2) minimize clinician burden, and (3) measure clinically relevant performance for health…
[17] 2026-09-21 Lessons Without Borders? Evaluating Cultural Alignment of LLMs Using Multilingual Story Moral Generation

arXiv:2604.08797v2 Announce Type: replace Abstract: Stories are key to transmitting values across cultures, but their interpretation varies across linguistic and cultural contexts. Thus, we introduce multilingual story moral generation as a novel culturally grounded evaluation task. Using a new dataset of human-written story morals collected across 14 language-culture pairs, we compare model outputs with human interpretations via semantic similarity, a…
[18] 2026-09-21 Cultural Alignment in Large Language Models Using Soft Prompt Tuning

arXiv:2503.16094v2 Announce Type: replace Abstract: Large Language Model (LLM) alignment is commonly achieved through supervised fine-tuning or reinforcement learning, both of which require labeled or preference data and update model weights. Without targeted cultural adaptation, however, deployed LLMs often exhibit culturally homogeneous behavior that fails to reflect diverse local values. Aligning models to cultural value frameworks such as Hofstede's Value…
[19] 2026-09-21 TabScope: Question-Adaptive Scope Selection for Table Question Answering

arXiv:2609.03395v2 Announce Type: replace Abstract: Large Language Models (LLMs) have shown strong performance on table question answering, yet their accuracy often degrades as table size increases. We find that this degradation is not uniform across question types. Localization-sensitive questions are particularly affected by irrelevant table content, while questions requiring broader evidence may still benefit from full-table reasoning. Based on this…
[20] 2026-09-21 Sixteen models, fewer than two voices: measuring ensemble dispersion where no answer is uniquely correct

arXiv:2608.00285v2 Announce Type: replace Abstract: Sixteen language models drawn from ten families produced, on average, the semantic diversity of 1.69 distinct formulations of a psychotherapeutic case, against a single-model baseline of 1.43 from one model's own runs. Ensembles place more than one reading before a decision-maker on the premise that several models supply several perspectives. Dispersion over their outputs is…
[21] 2026-09-21 Knowledge-Graph Based Augmentation versus Retrieval Augmented Generation for Cultural-Related Question Answering

arXiv:2609.18317v2 Announce Type: replace Abstract: Large language models (LLMs) suffer from a long-tail deficit: culturally specific facts, particularly those concerning underrepresented regions such as Latin America, appear too rarely in pretraining corpora to be reliably memorized. Retrieval-Augmented Generation (RAG) addresses this by grounding generation in external text, but structured alternatives such as Knowledge Graphs (KGs) offer tighter control over what…
[22] 2026-09-21 Data-Driven Integration Kernels for Interpretable Nonlocal Operator Learning

arXiv:2603.10305v4 Announce Type: replace Abstract: Machine learning models can represent climate processes that are nonlocal in horizontal space, height, and time, often by combining information across these dimensions in highly nonlinear ways. While this can improve predictive skill, it makes learned relationships difficult to interpret and prone to overfitting as the extent of nonlocal information grows. We address this challenge…
[23] 2026-09-21 ASGARD: Action-Space Guard for UAV Resilience via Reinforcement Learning

arXiv:2609.20982v1 Announce Type: new Abstract: Reinforcement learning (RL) controllers have been recently adopted for Unmanned Aerial Vehicles (UAV) navigation and control. However, they are susceptible to action-space attacks that overwrite the action commands after the policy generates them and before the actuators execute them. While most existing defenses target attacks on the policy's inputs, those addressing action-space attacks retrain the…
[24] 2026-09-21 OverThink: Slowdown Attacks on Reasoning LLMs

arXiv:2502.02542v5 Announce Type: replace Abstract: A reasoning language model (RLM) generates costly reasoning tokens, often hidden from the users, that help it excel at many tasks. Our Overthink attack targets RLM-based applications (such as chatbots or coding agents) that rely on external context by forcing these models to generate substantially more reasoning tokens while still producing contextually correct answers. An…
[25] 2026-09-21 Generative inversion for early ranking of competing geologic interpretations

arXiv:2609.20978v1 Announce Type: new Abstract: High-consequence subsurface decisions are often made under severe data scarcity. Experts may arrive at competing interpretations of the same subsurface system, yet early in a project there is rarely a practical way to determine which one is most realistic. This uncertainty can persist until several wells are drilled, often costing millions of dollars. Existing approaches…
[26] 2026-09-21 Beyond Benchmark Scores: Auditing Medical Vision-Language Models for Chest X-Ray Tuberculosis Screening

arXiv:2609.21763v1 Announce Type: cross Abstract: A medical model's benchmark score does not establish that the same conclusion holds under a different evaluation. This study tests whether claims about model ranking, score reliability and screening performance survive changes in cohort, prompt, negative spectrum, specified prevalence and operating threshold. We audit three medical vision-language models (BioMedCLIP, CheXficient, and MedSigLIP) and a general-domain…
[27] 2026-09-21 From Switching to Dynamic Regret: A Simple Reduction via Unbiased Random Sequences

arXiv:2609.20968v1 Announce Type: new Abstract: In non-stationary online learning, dynamic regret has attracted increasing attention as a measure of how well an online learner performs against a time-varying comparator sequence. Despite considerable advances, attaining optimal bounds for strongly convex and exp-concave losses often involves intricate analysis. In this paper, we present a textit{simple} framework that reduces dynamic regret minimization to…
[28] 2026-09-21 Predictive Suppression Layers for Communication-Efficient Spiking Neural Networks

arXiv:2609.21583v1 Announce Type: cross Abstract: Feedforward Spiking Neural Networks (SNNs) typically propagate every generated spike indiscriminately, disregarding whether the information is redundant from an information-theoretic perspective. This lack of selectivity induces high redundancy in inter-layer communication, creating an expensive overhead, e.g., in scenarios involving many-core neuromorphic hardware or communication-dominated Internet-of-Things (IoT) where features are transmitted wirelessly. To address this challenge,…
[29] 2026-09-21 Elastic Threshold Attention: Learned Contextual Sparsity for Long-Context Decoding

arXiv:2609.20888v1 Announce Type: new Abstract: Massive KV caches can cause severe memory-bandwidth bottlenecks during long-context decoding. Sparse attention methods mitigate this via selective loading, but that comes at a cost: rigid heuristics drop necessary context, leading to quality degradation. We introduce textbf{Elastic Threshold Attention (ETA)}, an end-to-end trainable architecture that achieves hardware-accelerated decoding speed without sacrificing dense model quality. ETA…
[30] 2026-09-21 Bio-MF: Low-Latency and High-Fidelity EEG-to-fNIRS Cross-Modal Generation for Hybrid Motor-Imagery Brain–Computer Interfaces

arXiv:2609.20904v1 Announce Type: new Abstract: Hybrid motor-imagery brain-computer interfaces (MI-BCIs) combining EEG and fNIRS can outperform EEG-only systems by exploiting complementary electrophysiological and hemodynamic information. To obtain such hybrid information when paired EEG-fNIRS acquisition is unavailable or inconvenient, recent studies have focused on EEG-to-fNIRS cross-modal generation. However, existing methods still suffer from slow generation and often require pretraining, limiting their…
[31] 2026-09-21 Continuous Delayed-Memory Stochastic Gradient Descent and Continuous-Time Reinforcement Learning from History of Astrophysical Time Series Studies

arXiv:2609.20906v1 Announce Type: new Abstract: Quasars are luminous objects in the universe that exhibit stochastic brightness variations encoding information about the supermassive black holes powering them, and modeling these variations from ground-based survey data time series, known as light curves, is a statistical challenge. This paper reviews how stochastic differential equations (SDEs) have been adapted with neural network parameterizations to…
[32] 2026-09-21 When AI Reviews Train AI Reviewers: Scientific-Judgment Collapse and Mitigation

arXiv:2609.20942v1 Announce Type: new Abstract: Large language models (LLMs) increasingly participate in scientific evaluation, both as automated reviewers and as assistants to human reviewers. As model-generated reviews enter public data and future training corpora, AI peer review can become recursive: later reviewers learn from judgments produced by earlier models. We study one step of this feedback loop in a controlled…
[33] 2026-09-21 Do Quantum Models Scale Like LLMs?

arXiv:2609.20912v1 Announce Type: new Abstract: In this work, we study the neural scaling laws of RydbergGPT, an autoregressive transformer model trained on qubit projective measurement data gathered from interacting Rydberg atom arrays. The quantum system is known to exhibit a finite-size remnant of a critical point as the laser detuning parameter is varied. We find that near the critical point…
[34] 2026-09-21 Riemannian Simultaneous Inference for Tangent Vector Field Regression

arXiv:2609.21910v1 Announce Type: cross Abstract: We consider nonparametric tangent vector field regression on a Riemannian manifold without boundary. Because responses at different points lie in different tangent spaces, the proposed kernel estimator first parallel transports nearby responses to the target tangent space and then forms a volume-corrected local average. We first derive its uniform second-order bias, finite-bandwidth covariance, and stochastic…
[35] 2026-09-21 Efficient Bayes-Adaptive Reinforcement Learning with Temporal Logic Specifications

arXiv:2609.20954v1 Announce Type: new Abstract: We present a novel end-to-end model-based Reinforcement Learning (RL) algorithm for efficient policy synthesis under given Linear Temporal Logic (LTL) specifications (e.g., safety or reachability) in unknown environments. To do so, a Limit-Deterministic B{"u}chi Automaton (LDBA) representation of the LTL task is synchronised with a Bayes-Adaptive Markov Decision Process (BAMDP) representation of the environment, which…
[36] 2026-09-21 Rhamba: Region-Aware Hybrid Attention-Mamba Framework for Self-Supervised Learning in Resting-State fMRI

arXiv:2605.01240v3 Announce Type: replace Abstract: Self-supervised pretraining is promising for large-scale neuroimaging, yet the impact of region-aware masking and hybrid sequence modeling remains underexplored. In this work, we introduce Rhamba, a region-aware pretraining framework that integrates anatomically guided masking with hybrid Attention-Mamba architectures for resting state functional magnetic resonance imaging (fMRI) analysis. Models were pretrained on the ABIDE dataset using…
[37] 2026-09-21 PRIVET: PRoximIty leakage detection Via Extreme value Theory

arXiv:2510.24233v2 Announce Type: replace Abstract: Deep generative models are often trained on sensitive data, such as genetic sequences, health data, or more broadly, any copyrighted, licensed or protected content. This raises critical concerns around privacy-preserving synthetic data, and more specifically around privacy leakage, an issue closely tied to overfitting. Existing proximity-based methods mostly assess privacy risk through global criteria, which…
[38] 2026-09-21 Large Language Models As Shannon Lossy Compressors Not Solomonoff Induction Estimators: The Singularity Is Not Near Without Symbolic Model Synthesis

arXiv:2601.05280v5 Announce Type: replace-cross Abstract: On the one hand, the question of whether Large Language Models (LLMs) are Solomonoff induction estimators has become an explicit question at the intersection of Algorithmic Information Theory (AIT) and Machine Learning (ML) of great interest. On the other hand, the now old idea of an AI Singularity that requires a reliable positive-feedback process in…
[39] 2026-09-21 Write on Paper and Get the Online Digital Trace: A New Era for Handwriting

arXiv:2609.12702v2 Announce Type: replace Abstract: Capturing the digital trace of handwriting usually requires a specific stylus and a compatible substrate, be it a capacitive touchscreen, an ElectroMagnetic Resonance (EMR) tablet as used in Wacom systems or special paper. While writing on regular paper offers rich haptics, no latency and is well known for improving information retention, no low-cost and widely…
[40] 2026-09-21 GVPO++: Group Variance Policy Optimization for LLM Post-Training and On-Policy Distillation

arXiv:2609.21432v1 Announce Type: cross Abstract: Post-training plays a pivotal role in enhancing the reasoning capabilities and task-specific expertise of large language models (LLMs). Despite recent advances in post-training methods, such as Group Relative Policy Optimization (GRPO), their practical deployment remains impeded by training instability arising from the reliance on importance sampling. We introduce Group Variance Policy Optimization (GVPO), a novel…
[41] 2026-09-21 The Binary Tree Mechanism is Optimal for Differentially Private Continual Counting

arXiv:2607.00876v3 Announce Type: replace-cross Abstract: Private continual counting is a fundamental problem in differential privacy: given a binary stream of length $n$, where each $1$ corresponds to the contribution of one individual, the goal is to release all running counts while protecting the privacy of each individual. For fixed privacy parameters, the standard binary tree mechanism achieves expected $ell_infty$ error…
[42] 2026-09-21 AI-GRACE: A Use-Case Operationalization Framework for Agentic AI: From Organizational Objectives and Obligations to Deployment Capabilities and Architecture

arXiv:2609.21192v1 Announce Type: new Abstract: Organizations deploying agentic artificial intelligence must determine more than whether a model is trustworthy; they must establish what to validate, control, and observe for a use case to deliver its intended outcome while meeting applicable obligations. This paper proposes AI-GRACE (Agentic Intelligence-Governance, Risk, Assurance, Controls, and Evidence) as a use-case operationalization framework connecting organizational governance…
[43] 2026-09-21 Generalizing Beyond Suboptimality: Offline Reinforcement Learning Learns Effective Scheduling through Random Solutions

arXiv:2509.10303v3 Announce Type: replace-cross Abstract: Online reinforcement learning (RL) approaches have demonstrated strong performance on Job Shop Scheduling (JSP) and Flexible JSP (FJSP) problems by learning scheduling policies through direct interaction with simulated environments. However, these methods often require extensive training interactions, limiting their sample efficiency and practical applicability. Motivated by this challenge, we introduce Conservative Discrete Quantile Actor-Critic (CDQAC),…
[44] 2026-09-21 Implicit Rule Induction with Test-Time Task Embeddings in ARC-like Tasks

arXiv:2609.21181v1 Announce Type: new Abstract: The Abstraction and Reasoning Corpus and related benchmarks evaluate whether AI models can solve novel reasoning tasks, but often leave unclear whether success reflects inference of the intended underlying rule or reliance on shortcuts. We address this gap by studying test-time task embeddings in Vision ARC (VARC), a model in which a pre-trained backbone is…
[45] 2026-09-21 Balance of Benchmarks: Semantic Density Reweighting for Task-Conditioned Model Comparison

arXiv:2608.30044v3 Announce Type: replace Abstract: Model comparison increasingly relies on large collections of publicly reported benchmark scores, yet common aggregation strategies trade off evidence coverage against control over capability weighting. Manually curated suites leave potentially informative evaluations unused, while uniform averaging retains them but gives greater influence to capabilities that happen to be benchmarked more densely. We introduce Balance of…
[46] 2026-09-21 SpecOpt: Contact-Diff Reasoning for Agentic Molecule Optimization Toward Binding Specificity

arXiv:2609.21165v1 Announce Type: new Abstract: Off-target protein binding is a major source of adverse effects for small-molecule drugs, yet most structure-based molecular design methods focus on generating selective compounds de novo rather than improving the selectivity of existing, well- characterized drugs. We introduce specificity optimization (SpecOpt), a molecular design task that seeks constrained structural modifications to an existing compound that…
[47] 2026-09-21 Beyond Final Answers: CRYSTAL Benchmark for Transparent Multimodal Reasoning Evaluation

arXiv:2603.13099v3 Announce Type: replace Abstract: We introduce CRYSTAL (Clear Reasoning via Yielded Steps, Traceability, and Logic), a diagnostic benchmark with 6,372 instances that evaluates multimodal reasoning through verifiable intermediate steps. We propose two complementary metrics: Match F1, which scores step-level precision and recall via semantic similarity matching, and Ordered Match F1, which further penalizes disordered reasoning chains. References are constructed…
[48] 2026-09-21 CaLR: Causal Latent Revision for Robust Diffusion Reasoning

arXiv:2609.20981v1 Announce Type: new Abstract: Autoregressive (AR) models suffer from local greediness, while diffusion language models (DLMs) often lack the strict causal structure required for reasoning. To combine the advantages and overcome the drawbacks of the dual, we propose Causal Latent Revision (CaLR), a framework that reformulates reasoning as constrained latent optimization. By adopting a causal topology matrix (CTM) from…
[49] 2026-09-21 LoRA Enhanced Contrastive Learning with SAS Vision Transformers

arXiv:2609.21061v1 Announce Type: new Abstract: Automatic target recognition (ATR) with synthetic aperture sonar (SAS) supports advanced naval capabilities, but deep learning is constrained by scarce target imagery, background clutter, and human-in-the-loop assessment. We adapt DINOv3 Vision Transformer (ViT) models to underwater SAS ATR using a three-stage parameter-efficient framework. Stage 1 uses Low-Rank Adaptation (LoRA) while freezing the ViT backbone, bridging…
[50] 2026-09-21 Detecting Hallucination in LLMs: Tracing the Topological Signatures of Impaired Context Sharing

arXiv:2609.21096v1 Announce Type: new Abstract: In this work, we examine the topology of information flow patterns within attention graphs to effectively distinguish hallucinated from non-hallucinated responses. We analyze the Forman-Ricci curvature to identify structural patterns indicating information bottlenecks in attention graphs. We then introduce a method that captures both semi-local and global information-flow characteristics of attention heads associated with hallucinated…
[51] 2026-09-21 TinyCeNN-LM: Quality-Gated Conversion of Pretrained Attention with CeNN-Inspired Cellular-Recurrent Layers

arXiv:2609.21139v1 Announce Type: new Abstract: Replacing attention in a pretrained language model is a compatibility problem: a plausible substitute may alter representations expected by later layers. TinyCeNN-LM introduces a emph{quality-gated post-training conversion} framework using CeNN-inspired cellular-recurrent layers with bounded local processing, compact recurrent memory, routing, fusion, and accept-or-rollback validation. Three implementations are studied: Integrated Memory, MemoryFusion, and PDelta3-GDN2-CLVR+Local32. Strict PDelta3…
[52] 2026-09-21 Decoupling Internal Representational Changes and Causal Importance in Fine-Tuned Large Language Models

arXiv:2609.21113v1 Announce Type: new Abstract: Fine-tuning has emerged as a widely adopted approach for adapting LLMs to a variety of downstream tasks. However, how it reshapes their internal mechanisms remains poorly understood. To address this, we investigate how fine-tuning alters internal representations in LLMs, including attention patterns and layer-wise activations, and examine whether these changes are linked to task-relevant components…
[53] 2026-09-21 Can Agents Design Better Chips with a Higher Level Abstraction?

arXiv:2609.21157v1 Announce Type: new Abstract: Large Language Model (LLM) agents are increasingly being explored for chip design, but most existing approaches operate directly at RTL. We ask whether agents can design better chips by leveraging higher-level abstractions. We compare Direct RTL Design, Agent-based HLS Design, Post-Compiler HLS Refinement, and Post-HLS RTL Refinement, and combine Agent-based HLS Design with Post-HLS RTL…
[54] 2026-09-21 Clinician-Grounded Quality Assurance for AI-Assisted Psychiatric Intake

arXiv:2609.21149v1 Announce Type: new Abstract: Before patients can use AI-assisted psychiatric intake systems, health systems need practical ways to routinely evaluate these tools against their clinical standards for quality assurance. Because clinicians may use different intake styles, evaluation for this task must (1) support comparison across interviewing approaches, (2) minimize clinician burden, and (3) measure clinically relevant performance for health…
[55] 2026-09-21 HERMES: A Holistic End-to-End Risk-Aware Multimodal Embodied System with Vision-Language Models for Long-Tail Autonomous Driving

arXiv:2602.00993v2 Announce Type: replace-cross Abstract: End-to-end autonomous driving models increasingly benefit from large vision-language models for semantic understanding, yet safe and reliable planning under long-tail conditions remains challenging, particularly in mixed-traffic environments involving heterogeneous road users and rare safety-critical interactions. This paper proposes HERMES, a holistic risk-aware end-to-end multimodal driving framework that explicitly incorporates long-tail semantic knowledge into trajectory planning.…
[56] 2026-09-21 A Unified Evaluation Framework for Trustworthy Large Language Models, Agentic AI, and Multimodal Systems

arXiv:2609.19524v2 Announce Type: replace Abstract: Benchmark scores alone provide an incomplete basis for assessing the trustworthiness of modern artificial intelligence systems. Large language models (LLMs), agentic systems, and multimodal models (MLLMs) require different forms of assessment, yet their evaluation evidence must remain interpretable for development and oversight. We propose a unified framework that connects output-level, trajectory-level, and cross-modal assessment through…
[57] 2026-09-21 Multi-turn Conversational AI from Text to Multimodal Interaction: Data, Models, Evaluation, and Open Challenges

arXiv:2608.17605v2 Announce Type: replace-cross Abstract: Conversational AI is moving beyond isolated text prompts toward sustained, multimodal interaction. In real conversations, users clarify goals, revise requests, interrupt responses, switch topics, and introduce new evidence while expecting systems to preserve context across turns. This makes multi-turn dialogue a distinct challenge requiring systems to maintain and update memory, ground responses across modalities, tools,…
[58] 2026-09-21 LiteMedCoT-VL: Parameter-Efficient Adaptation for Medical Visual Question Answering

arXiv:2605.09384v2 Announce Type: replace-cross Abstract: The reasoning gap between large and compact vision-language models (VLMs) limits the deployment of medical AI on portable clinical devices. Compact VLMs of 2-4B parameters can run on resource-constrained hardware but lack the multi-step reasoning capacity needed for interpretable clinical decision support. Existing knowledge distillation methods transfer answers without the reasoning process behind them. Medical…
[59] 2026-09-21 Matrix AdaGrad: Row-wise and Column-wise Adaptive Subgradient Methods

arXiv:2609.21815v1 Announce Type: cross Abstract: Adaptive optimization methods such as AdaGrad and Adam are widely used in modern neural-network training, but their adaptive scaling is primarily designed for vector-valued parameters and does not explicitly exploit matrix structure. Recent matrix-aware optimizers demonstrate the benefits of structured optimization, yet a general theoretical framework for deriving matrix-aware adaptivity comparable to that of AdaGrad…
[60] 2026-09-21 CPR: Combining global composing, local performing and full-sequence refining in piano rendering with continuous autoregressive modelling

arXiv:2609.18216v2 Announce Type: replace-cross Abstract: Prompt-conditioned piano MIDI-to-Music rendering aims to faithfully render target notes while reproducing the timbre of a reference recording. Existing approaches primarily follow two paradigms: autoregressive (AR) modeling and flow matching (or diffusion). Discrete-codec AR models provide causal temporal modeling, but quantization can discard acoustic detail. Flow matching better preserves acoustic structure in the cost of…
[61] 2026-09-21 NemotronLabs VoiceChat: An Open Full-duplex Speech-to-Speech Model with Tool Calling Capabilities

arXiv:2609.21967v1 Announce Type: cross Abstract: We introduce NemotronLabs VoiceChat, an open full-duplex speech-to-speech model with native tool-calling capabilities. NemotronLabs VoiceChat combines a streaming speech encoder and decoder-only language model with parallel specialized output streams for agent text and structured function calls, an auxiliary RNN-T branch for incremental user transcription, and a streaming TTS decoder. This design enables the model to…

Where Kimbodo Comes In

Kimbodo builds and operates this in production for businesses — see our AI Consulting & Strategy practice, or Request an AI Roadmap.

AI Research & Papers — September 21, 2026

Findings

Where Kimbodo Comes In

Sources

Chad Collins

Leave a comment Cancel reply

You May Also Like

How to Turn This Month’s Multimodal and Agentic AI Research into Safer, Higher‑value Production Systems

What New AI Research Changes How Businesses Build Auditable, Robust and Efficient Production AI

Engineering Founded & Customer Focused