Digital Libraries
Zero-Cost Detection of Fabricated Citations: A Staged Heuristic Approach for Computing Dissertation Assessment
Hazrat Ali and Mohamed Elawady
Generative AI's growing role in academic writing has driven a rise in fabricated references across dissertations and student reports. We present the Fabrication Severity Stack (FSS), a three-tiered framework classifying citation fabrications by detection complexity - from non-existent references to subtle content misattribution. Building on the FSS, we propose a staged heuristic verification workflow, executable through mainstream LLM interfaces. We argue that this framework offers computer science educators/examiners a practical, zero-cost basis for
embedding reference verification into the assessment of dissertation and students' reports. In a pilot study on 100 citations (20% seeded with Level 1 and Level 2 fabrications), the proposed heuristic workflow enabled mainstream LLM interfaces to correctly flag 9-10 of 10 non-existent references and partially detect metadata corruption. These preliminary results suggest that freely accessible, general-purpose LLMs can achieve high recall on citation verification without reliance on proprietary detection tools.
Read the full abstract ▾
Digital Libraries
Building Knowledge Graphs for Theses: A Survey
Donia Ruelas, Elmer Coyla, Miguel Aceituno et al. (+1 more)
Academic theses are a valuable source of scientific knowledge, yet their semantic exploitation remains limited. While Scholarly Knowledge Graphs have significantly advanced the representation of scientific publications, knowledge graph construction for academic theses has received comparatively limited
attention. This paper provides a systematic review of existing approaches and analyzes their potential application for the thesis domain. We analyzed 46 peer-reviewed studies published between 2021 and 2025, and organised existing approaches into a structured classification of techniques, models, and tools. The review reveals a methodological
evolution from natural
language processing techniques for entity and relation extraction to ontology engineering, Semantic Web technologies, and AI-driven enrichment through knowledge graph
embeddings, Linked Open Data, and Large
Language Models. The findings provide a consolidated view of the current state of the art and identify methodological foundations and future research directions for building knowledge graphs for the thesis domain.
Read the full abstract ▾
All authors Donia Ruelas, Elmer Coyla, Miguel Aceituno, Elwin Huaman
Artificial Intelligence
Doing Less with More: A First-Principles Exploration of the Suitability of Agentic Computing over Alternative Architectural Choices
Ritvik Garimella, Biplav Srivastava and Amit Sheth
There is growing interest in automating business activities with Agentic Artificial Intelligence (AI) due to latter's seeming ease of use. Never has it been easier, or costlier, to do less with more. However, little is known about when
agents are preferable to established alternatives such as local computation, Representational State Transfer (REST), the Simple Object Access Protocol (SOAP), and the Model Context Protocol (MCP), particularly when development speed, performance, and operational cost are considered. We investigate this question using a controlled mathematical task that compares seven methods on a
benchmark of 1,000 arithmetic expressions where semantics of operator precedence has to be preserved. We ran this setup across a native Function Calling baseline (Python library), REST, SOAP and MCP microservice architectures, direct responses from large
language models (LLMs) using a Groq cloud API and five locally deployed Ollama models, and agentic systems in which five local LLMs invoke MCP tools. The deterministic methods achieve 100% accuracy with negligible latency. Direct LLM evaluation achieves at most 92.9% accuracy at over 22,000x the latency of a native function call. With MCP tools, the best
agent reaches 99.9% accuracy, but incurs nearly 1,000,000x the baseline latency. These results demonstrate that LLMs approaches are poorly suited for exact arithmetic evaluation and motivate an informed architectural decision framework (ADF) for choosing an architecture based on accuracy and throughput requirements. We recognize that an actual interaction with an
agent may span both
reasoning and linguistic tasks. Hence our recommendation is to explore multiple architectural choices (based on task type) rather than one-size-fits-all solution (LLMs). Code is available at - https://github.com/Ritvik-G/adf/
Read the full abstract ▾
Computational Geometry · Cross-listed from math.CO
Almost Empty Monochromatic Polygons in Planar Point Sets
Bhaswar B. Bhattacharya, Sandip Das, Sk Samim Islam et al. (+2 more)
For integers $r \ge 3$, $c \ge 2$, and $s \ge 0$, let $\mathsf{M}_r^{\star}(c,s)$ denote the smallest integer such that any set of at least $\mathsf{M}_r^{\star}(c,s)$ points in the plane in general position (no three on a line), colored with $c$ colors, contains a monochromatic $r$-gon (not necessarily convex) with at most $s$ interior points. Define $\lambda_r^{\star}(c)$ to be the smallest integer such that $\mathsf{M}_r^{\star}(c,\lambda_r^{\star}(c)) < \infty$. The case of triangles, that is, $r=3$, has been studied in recent years. In this paper, we establish bounds on $\lambda_r^{\star}(c)$, for $r \ge 4$. Specifically, in the bichromatic case, we prove that $\lambda_r^{\star}(2)\leq \lfloor\frac{r-2}{2}\rfloor$, for every $r\geq3$. For hexagons, we sharpen this bound to $\lambda_6^{\star}(2)\leq1$, that is, every sufficiently large bichromatic planar point set in general position contains a monochromatic hexagon with at most one interior point. Next, for any number of colors $c\geq3$ and $r\geq3$, we show that $$(r - 2) \left \lfloor \frac{c-1}{2} \right \rfloor \leq \lambda_r^\star(c) \leq (r-2)(c-2)+\left\lfloor(r-2)\delta_c\right\rfloor, $$ for some constant $\delta_c \in (0, \frac{1}{2}]$, depending on $c$. The upper bounds combine spiral polygonizations with nested convex-hull arguments and, in the hexagon case, incidence counting in $4$-connected triangulations. The lower bounds follow from colored Horton-set constructions.
Read the full abstract ▾
All authors Bhaswar B. Bhattacharya, Sandip Das, Sk Samim Islam, Aashirwad Mohapatra, Saumya Sen
Computer Vision
Vocabulary-Incomplete Unsupervised Vision-Language Model Adaptation
Si Qin, Jiawei Tang, Yaxin Hou et al. (+4 more)
In certain real-world scenarios, obtaining labeled data is excessively expensive, motivating increasing interest in adapting models without annotations.
Vision-
language models (VLMs) provide pseudo-labels for unlabeled images through image-text similarity, enabling adaptation without manual labeling. The previous unsupervised
vision-
language model adaption methods assume access to an adaptation vocabulary containing all class names. However, a complete vocabulary is difficult to construct in real-world scenarios, as unlabeled data may contain classes that are overlooked or fall outside the current task scope. Thus, we propose Vocabulary-Incomplete Unsupervised
Vision-
language model Adaptation (VI-UVA), where only a subset of class names are available during adaptation. The vocabulary incompleteness induces an adaptation bias toward provided classes, which degrades omitted-class recognition. To address this issue, we propose Label Space Reconstruction (LSR), which reconstructs a latent label space from the unlabeled data and incorporates it into
prompt tuning, thereby extending adaptation beyond the incomplete vocabulary and mitigating the adaptation bias. Furthermore, we derive an omitted-class semantic discrepancy bound for LSR and identify the key factors affecting the bound. Extensive experiments across eight
benchmarks demonstrate that LSR outperforms existing state-of-the-art methods.
Read the full abstract ▾
All authors Si Qin, Jiawei Tang, Yaxin Hou, Yan Chen, Yutong Xie, Chen Li, Yuheng Jia
Computer Vision
Cross-Modal Forgetting and Transfer in Unified Multimodal Post-Training
Francesco Colasurdo, Ivan Slobozhan, Joaquin Vanschoren et al. (+1 more)
Unified
multimodal models answer questions about images and also generate images with one shared set of weights. This creates a post-training risk that single-purpose models do not have: fine-tuning one capability can damage another that works in a different modality. We measure this cross-modal forgetting for LoRA post-training of Janus-Pro-7B. Every target is chosen by how poorly the base model already handles it, from a poetic register to a
language it barely speaks, so that retention is tested against a real intervention; we then evaluate the capabilities we never trained, first after a single run and then after chaining three. A single run looks entirely safe: nothing we did not train degrades on four public
benchmarks or in free-form English, even when the adapter cuts held-out Welsh perplexity by 76% with a relative weight change under 2%. That safety belongs to the run, not to the recipe. Chaining three runs destroys the capability learned one run earlier and raises general-text perplexity by one to two orders of magnitude above the never-adapted model, on the very measures that stayed flat before. The cause is in the objective, which constrains only the modality being trained and leaves the other head unconstrained: cutting the learning rate to one fifth does not prevent the collapse, while anchoring the other head removes the damage to the capabilities we never trained, at no cost to what is learned. What is learned also stays inside its own modality, with Welsh acquired from text closing only 3% of the English-Welsh gap in image generation.
Read the full abstract ▾
All authors Francesco Colasurdo, Ivan Slobozhan, Joaquin Vanschoren, Anna Vettoruzzo
Software Engineering
Eliciting Specifications: When Does Interactive Clarification Help?
Mowafak Allaham, Aditya Vempaty, Ashish Jagmohan et al. (+2 more)
LLMs can generate Lean specifications from task descriptions, but those descriptions are often underspecified or ambiguous and may omit verifier-relevant edge cases, yielding imperfect formal specifications. We study whether an elicitation
agent (EA) can improve verifier-checked Lean specifications by seeking clarification from a reference-grounded simulated developer before formalization. The
agent decomposes the task, identifies unresolved decisions, asks bounded clarification questions, and uses the answers to write a refined natural-
language specification that an LLM formalizer then translates into Lean. On 189 Verina Lean 4 specification-generation tasks, we compare single-shot generation with elicitation under question budgets of k=3 and k=10, allowing at most three or ten EA questions per task, across same-model and cross-model configurations. The same-model Claude Sonnet 4.6 pipeline has negative strict pass@1 estimates at both budgets, showing elicitation is not a uniform accuracy booster. In contrast, the Sonnet 4.6-to-GPT-5.6 and Sonnet 4.6-to-Gemini 3.8 Flash pipelines have positive strict estimates. The largest strict result reaches 65.6%, a +6.3-point change (+12 tasks), in the Sonnet 4.6-to-GPT-5.6 pipeline. Six of eight conditions have positive compile-conditional estimates, with a maximum cross-model difference of +4.9 points among compiling outputs. Treatment estimates also differ across the basic and advanced splits; the largest advanced-task change occurs in the Sonnet 4.6-to-GPT-5.6 configuration at k=10, from 42.0% to 50.6%. Failure analysis identifies postconditions as the largest source of degradation. These findings suggest future autoformalization systems should treat elicited natural-
language specifications as intermediate artifacts and optimize jointly for semantic correctness among compiling outputs and reliable translation of clarified information into Lean.
Read the full abstract ▾
All authors Mowafak Allaham, Aditya Vempaty, Ashish Jagmohan, Prasenjit Dey, Ravi Kokku
Computational Geometry
New certified lower bounds for the moving sofa problem in a three-dimensional snaky corridor
Yuzhe Tang
Consider the "snaky" corridor $K_4 \subset \mathbb{R}^3$ from Problem 6.64 of arXiv:2511.02864: a tube of unit-square cross-section whose two right-angle turns lie in perpendicular planes, with a straight middle section of clear length $4$. We exhibit explicit rigid motions of connected bodies through $K_4$ and certify their volumes with the problem's own evaluation pipeline, unmodified, at the tight motion tolerances of its final scoring stage (translation $0.005$, rotation $0.05$ degrees). The certificates give $V^*(K_4) \ge 1.8355162$ at grid fineness $0.002$ and $V^*(K_4) \ge 1.8558404$ at grid fineness $0.001$, against the previously reported $1.81$; rerunning the pipeline's published best path under the identical configuration yields $1.8179428$, so every comparison is path against path under one fixed certifier. Three stages produce the record motion: a simple "staggered twist" backbone in which the whole 90-degree twist about the tube axis happens inside the second turn; evolutionary (CMA-ES) refinement, including a mixed-fidelity scheme that re-scores each generation's leading candidates on a grid twice as fine and lets those ranks steer recombination; and a gradient polish carried out directly at the finer carve resolution, guarded by periodic rebuilding of the frozen interpolation structure and by a penalty on control-pose spacing. These guards close three documented failure modes in which an optimizer inflates its own objective: frozen-structure inflation, spacing stretch, and exploitation of a coarse grid by optimizing and selecting on it. A staircase-based predictor with two grid bases, anchored by matching grid sensitivity, forecast the certificates to about $0.001$ before the runs. With the companion note's upper bound $V^*(K_4) \le 3.4083$, the problem now sits in the bracket $1.8558404 \le V^*(K_4) \le 3.4083$. All paths, notebooks and certified results are public.
Read the full abstract ▾
Computational Geometry
An upper bound for the three-dimensional moving sofa problem in a snaky corridor
Yuzhe Tang
Let $K_M \subset \mathbb{R}^3$ be the "snaky" corridor with unit-square cross-section: an infinite entry arm parallel to the $x$-axis, a middle arm parallel to the $y$-axis whose clear straight section has length $M$, and an infinite exit arm parallel to the $z$-axis, with the two turns lying in the $x$-$y$ and $y$-$z$ planes. Let $V^*(K_M)$ be the supremum of the volumes of connected bodies that can be moved continuously (by rigid motions) from the entry arm to the exit arm while remaining inside $K_M$. We prove $V^*(K_M) \le 7.8691$ for every $M \ge 3.53$ (
Theorem A); by a sharper analysis of the same mechanism, $V^*(K_M) \le 4.2862$ for every $M \ge 2.61$ (
Theorem B); and, by replacing the final analytic caps with a certified branch-and-bound computation over the compact family of entry tubes compatible with the forced events, $V^*(K_M) \le 3.4083$ for every $M > 2.2$ (
Theorem C; computer-assisted, in double precision with conservative slack). In particular all three bounds apply to the corridor with $M = 4$ used in Problem 6.64 of Georgiev, Gómez-Serrano, Tao and Wagner (arXiv:2511.02864), and uniformly to every "turns far apart" reading of that problem. To the best of our knowledge, this is the first upper bound for the snaky-corridor 3D moving sofa problem; previously not even finiteness of $V^*$ had been recorded. This note proves upper bounds only; combined with the certified lower bound 1.8558404 from the author's separate certification work via the problem's own evaluation pipeline, the state of the problem becomes $1.8558404 \le V^*(K_4) \le 3.4083$.
Theorems A and B are elementary and self-contained: a first-crossing (elbow) event forces a definite tilt of the body's own entry tube, and Fubini-type chord estimates for tilted tubes cap the volume.
Theorem C certifies, by the same forced-event skeleton, the largest corridor volume that any admissible entry tube can carry.
Read the full abstract ▾
Programming Languages
Proceedings of the Seventh miniKanren and Relational Programming Workshop (miniKanren 2025)
Michael D. Adams and William E. Byrd
The miniKanren and Relational Programming Workshop brings together researchers, implementors, and users of the miniKanren family of relational programming
languages, including miniKanren, microKanren, core.logic, OCanren, Guanxi, and related systems. The workshop focuses on the design, implementation, and application of miniKanren-like
languages, as well as new techniques and tools for relational programming. These proceedings collect the eleven accepted papers presented at the Seventh miniKanren and Relational Programming Workshop.
Read the full abstract ▾
Hardware Architecture
CacheFit: The Rising Cost of Cache Replacement and Constraint-Driven Redesign of Last-Level Cache Replacement Policies
Bita Aslrousta, Kaushal Mhapsekar, Darsh Asher et al. (+3 more)
The latency, power, and area costs of LLC replacement remain poorly explored not because simulators cannot be instrumented–event counters feeding a power model are routine–but because no cost model takes a replacement policy as input. CACTI and McPAT price regular structures, a cache or a core, not the small, oddly shaped arrays and dependent logic a policy is made of, and no event count reveals a victim-selection path's logic depth. Without a model, cost has never been an objective a policy search could optimize. In practice, a policy is first selected in simulation for its IPC, and its cost is reduced later in RTL, after the
algorithm is fixed. Cost thus enters the design too late to shape it. This paper asks what happens when cost is present from the start. We build CacheFit, the first policy-design loop for LLC replacement with
circuit-level cost inside the search objective. CacheFit connects ChampSim to HARCOM, a
circuit-level cost model, and to a large
language model that proposes policy variants. Every candidate is scored on latency, area, power, and speedup while its
algorithm is still being written. The designer sets the area, power, and latency budgets, and every proposal and every lesson the search learns is stored as readable text that the designer can check and edit. We apply CacheFit to two published policies. From MPPPB, the most expensive baseline in power and transistors, it produces Fit-MPPPB: 60.3% less storage, 56.3% less static power, and 58.9% fewer transistors, with 1.1% higher IPC and a hit rate 7.5 percentage points higher on SPEC CPU2017. From SHiP++, the highest-IPC baseline, it produces Fit-SHiP++: 56% less storage, 61% less static power, and 56% fewer transistors, for 1.54% lower IPC and a hit rate 2.1 points lower. A policy designed with its cost in view can therefore be far cheaper to build at similar or better performance.
Read the full abstract ▾
All authors Bita Aslrousta, Kaushal Mhapsekar, Darsh Asher, Azam Ghanbari, Joshua Kalyanapu, Samira Mirbagher Ajorpaz
Machine Learning · Cross-listed from eess.IV
Geometry-Aware Diffusion Approximate Posterior Sampling for Sparse-View and Limited-Angle CT
Honglei Brinkmann, Carola-Bibiane Schönlieb and Ander Biguri
Sparse-view computed
tomography (CT) reduces radiation dose and acquisition time and may mitigate motion artifacts. However, angular undersampling provides insufficient information to determine the image uniquely and stably. Limited-angle CT, arising from restricted angular coverage, produces strongly directional information loss associated with the missing angular range. In both settings, image directions may be strongly observed, weakly constrained, or unobservable, leading to severe ill-posedness and reconstruction ambiguity. Existing
diffusion-based approaches incorporate measurement information through likelihood guidance, data-consistency operations, or range-null-space corrections. However, they do not generally use the continuously varying measurement sensitivity of the acquisition to jointly shape both reconstruction updates and
stochastic exploration. We propose a geometry-aware
diffusion-guided
stochastic reconstruction framework for sparse-view and limited-angle CT. Its central component is a regularized noise-weighted pullback metric constructed from the CT forward operator and measurement-noise covariance. This metric continuously adapts both the measurement-aware update and
stochastic exploration according to directional measurement sensitivity, suppressing changes along strongly constrained directions while permitting greater exploration along weakly constrained and unobservable directions. We complement this geometry-aware update with a regularized data-consistency correction and approximately null-space-restricted
stochastic perturbations, implemented
matrix-free using forward and backprojection operations together with conjugate-gradient solves. Experiments on sparse-view, noisy, and limited-angle CT demonstrate competitive reconstruction quality, strong measurement consistency, and spatially resolved empirical uncertainty estimates.
Read the full abstract ▾
Cryptography and Security
Adversarial RL for Port-Scan Evasion: Attacker Feature Visibility in Edge-Deployed IDS
Logan Andrew North, Priya Sanjay Kaluskar, Shasi Kumar Ramachandran Prabhu et al. (+2 more)
Machine learning-based intrusion detection systems (IDS) are increasingly used in resource-constrained Internet of Things (IoT) environments, yet their robustness is often evaluated against static attacks rather than adversaries that adapt to detection feedback. This paper investigates adaptive port-scan evasion against ML-based IDS models deployed on a Raspberry Pi 3B+. We implement a live Zeek-based IDS pipeline with XGBoost, a multi-layer perceptron, and a 1D convolutional neural network trained on TON_IoT telemetry, and use a Deep Q-Network (DQN) adversary to learn evasive combinations of probe timing, TCP flags, and payload size under black-box, gray-box, and white-box feature-visibility settings. Although the deployed IDS models detect conventional port scans at 91.1–99.8%, DQN final-50-episode evasion rates range from 61.9% to 98.3% across feature-visibility settings. Greater feature visibility does not monotonically improve evasion, and its effect is model-dependent: against XGBoost, the black-box
agent achieves 92.9% evasion, compared with 61.9% and 76.9% for gray-box and white-box
agents, respectively, whereas 1D-CNN is most vulnerable under white-box access at 98.1%. Because standard DQN can overestimate action values, we additionally spot-check representative conditions using Double DQN. The gray-box condition remains unstable in this check, providing no evidence that overestimation bias alone explains the observed instability. These results show that limited feature knowledge can still enable effective adaptive evasion against static edge-deployed IDS models, motivating more robust defenses for IoT edge environments.
Read the full abstract ▾
All authors Logan Andrew North, Priya Sanjay Kaluskar, Shasi Kumar Ramachandran Prabhu, Peilong Li, Suman Saha
Robotics
Trajectory Planning without Trajectory Data: A Manifold-Guided Approach
Silong Yong, Anji Liu, Cunxi Dai et al. (+5 more)
A common way for trajectory planning is to leverage generative models trained on large collections of expert trajectories. At inference time, the model generates executable trajectories by conditioning on task goal constraints. However, trajectory-based methods rely on costly supervision, scale poorly with sequence length, and often generalize poorly to unseen constraints such as novel start-goal pairs. We propose an alternative to learn the underlying state-space
manifold and use the geometry of the
manifold for trajectory planning. This approach requires only state observations and enables generalization to unseen constraints by con- structing trajectories on the learned
manifold of the state space. Experiments on Maze2D and robotic motion-planning
benchmarks show that Ariadne constructs feasible paths from state-only supervision and generalizes to unseen start-goal combinations. On high-dimensional dual-arm planning, it remains competitive with trajectory-supervised and classical planners, while requiring no trajectory data for training.
Read the full abstract ▾
All authors Silong Yong, Anji Liu, Cunxi Dai, Carl Busart, Guanya Shi, Yilun Du, Katia Sycara, Yaqi Xie
Robotics
CIRRA: Dual-Level Continual Instruction Reconciliation with Ongoing Execution for Embodied Robot Agents in Interactive Household Tasks
Ci Zhang, Enfu Nan, Arman Akbari et al. (+6 more)
Household
robots must accommodate new user instructions while executing ongoing tasks. Existing
agents often regenerate or extensively revise the remaining task sequence, introducing plan ambiguity, logical inconsistency, and redundant execution. We formulate continual instruction reconciliation and propose CIRRA (Continual Instruction Reconciliation for
Robot Agents), a dual-level framework combining LLM-based semantic
reasoning with rule-constrained structural integration. CIRRA first grounds incoming instructions to unique executable skills and resolves underspecified actions and execution locations. It then preserves the ongoing subtask sequence as an execution backbone and generates integration candidates by inserting incoming subtasks into location-matched segments. The semantic reasoner evaluates only modified segments to identify dependencies and conflicts and select the most logically coherent local integration. This structure-preserving process maintains
alignment with ongoing execution, mitigates ambiguity and inconsistency, and reuses shared subtasks to reduce redundant execution. We also introduce CHIRP (Continual Household Instruction Reconciliation and Planning), a text-based
benchmark of 120 episodes across eight household environments and six categories of everyday activities. On CHIRP, CIRRA achieves 74.2% decision agreement, exceeding the strongest replanning baseline by 30 percentage points; every correct fusion decision yields a correctly placed, conflict-free schedule. On a Unitree G1
humanoid, CIRRA interrupts ongoing skills at the correct moment in every trial and significantly outperforms all baselines on every metric.
Read the full abstract ▾
All authors Ci Zhang, Enfu Nan, Arman Akbari, Lin Zhao, Li Wang, Chen Wang, Weiwei Chen, Yanzhi Wang, Geng Yuan
Information Theory
Reconstruction of Geometric and Structural Aspects of Undeciphered Messages by Information-theoretic Methods
Hector Zenil and Abicumaran Uthamacumaran
In previous work, we established sufficient conditions for recovering geometry and
topology from non-random information, demonstrating reconstruction with the Arecibo message. Undeciphered historical records present a related problem: one-way communication without access to their makers' intentions or encoding conventions. Here we extend this framework to the Phaistos Disc, Andean khipu, Rongorongo, Indus inscriptions and the Voynich manuscript, using Egyptian layouts and Arecibo as controls. Algorithmic information dynamics (AID) guides structural perturbations, combining classical information measures, predictive coding, compression and algorithmic-probability-based estimates. Phaistos retains 0.714 bits of excess adjacent-sign mutual information after positional controls; its authentic
segmentation lies within a broad information basin, and cross-face boundary prediction achieves AUCs of 0.76 and 0.86. Repetition explains its principal recurrence, while the tested spiral
embeddings provide no evidence of additional cross-winding structure. Across several corpora, comparative transfer persists beyond exact shared local pairs. Khipu
knot forms distinguish authentic attachments after controlling for degree, depth, fibre and twist (adjusted $p=0.009$); constrained assignment recovers 8.75% of concealed parents against 7.87% expected by chance. Row-
alignment tests yield no comparable evidence after correction. Calibrated BDM recovers Arecibo's 23-column width (search-corrected $p=0.005$), while executable controls detect spatial dependence and eliminate it through interventions on the generator. These findings connect information-based structural reconstruction with interventional
reasoning, distinguishing recoverable organisation from semantic decipherment.
Read the full abstract ▾
Computation and Language
LRCC: Generalizing Low-Rank Compression with Conditional Computation
Thomas Vaitses Fontanari, Maximo Eduardo Rulli, Federico Alvetreti et al. (+2 more)
Low-rank compression reduces the cost of pretrained
language models by replacing linear transformations with low-rank factorizations. However, conventional methods use a fixed rank allocation during inference, assigning the same amount of compute regardless of the input
token. We introduce Low-Rank Conditional Computation (LRCC), which adds
token-dependent computation to pretrained models by training one lightweight router per
Transformer block to select among a small set of nested low-rank paths. During training, the low-rank factors remain frozen, and only the routers are optimized. We evaluate LRCC on Llama and Qwen models for
language modeling and zero-shot downstream tasks. Within the same average active-parameter budget, LRCC improves the predictive performance over static low-rank compression, including a 7.6 percentage-point gain in average downstream accuracy on Llama-2-7B over static methods. At matched batch-size-1 decoding latency, LRCC improves both perplexity and downstream accuracy on Llama-3.2-1B and remains competitive on Llama-2-7B, without specialized kernels. Finally, we assess the usefulness of assigning a
token-wise path by analyzing the routers' path choices.
Read the full abstract ▾
All authors Thomas Vaitses Fontanari, Maximo Eduardo Rulli, Federico Alvetreti, Donatella Genovese, Simone Scardapane
Machine Learning
Climbing the Design Ladder: Sequential Knowledge Distillation for Early-Stage Circuit Timing Prediction
Reza Moravej, Fahad Rahman Amik, Zhanguang Zhang et al. (+2 more)
Integrated
circuit design involves multiple design stages: logic synthesis, floorplanning, placement, and routing, with each stage taking hours to weeks to complete. Discovering timing violations late in this flow forces costly iterations back to earlier stages, wasting computational resources and delaying product launches. While predicting post-routing timing from early-stage data could prevent these failures, existing machine learning approaches struggle with the massive abstraction gap between post-synthesis logical descriptions and post-routing physical layouts. We propose STEP-KD (Sequential Timing Evaluation via Progressive Knowledge
Distillation), which leverages intermediate design stages as “stepping stones” for progressive knowledge transfer rather than attempting direct prediction. STEP-KD trains teacher models at the post-routing, post-placement, and post-floorplan stages, then sequentially distills their knowledge to a post-synthesis student model through representation
alignment. Experiments on diverse
circuits demonstrate that STEP-KD reduces timing prediction error compared to direct
distillation and supervised baselines, and in most settings compared to the industry-standard Static Timing Analysis (STA) tool. STEP-KD reduces the weighted mean absolute percentage error of Total Negative Slack prediction to 19.78%, compared with 74.84% for STA. Our proposed method is step forward to identify timing problems earlier, avoiding expensive late-stage redesigns.
Read the full abstract ▾
All authors Reza Moravej, Fahad Rahman Amik, Zhanguang Zhang, Didier Chételat, Yingxue Zhang
Robotics
HexaGripper: A Single-Actuator, Winch-Deployed Gripper for Autonomous Aerial Parcel Collection
Nimantha Adikaram, Mahen Abeyratne, Lakmina Chandrajith et al. (+3 more)
Autonomous aerial parcel collection remains constrained by package-transfer operations that require manual loading, landing, or dedicated ground infrastructure. This letter presents HexaGripper, a single-actuator, winch-deployed six-plate gripper for autonomous collection of cuboid parcels. The proposed mechanism provides synchronized grasping with a wide capture region and tolerance to positional misalignment during pickup. The system integrates
vision-based
alignment, range sensing, winch deployment, and state-based control to enable autonomous collection while maintaining
UAV separation from the pickup surface. Experimental evaluation covered static grasping, capture-workspace assessment, and end-to-end outdoor aerial collection. Static experiments achieved 100% pickup success (15/15 trials) across three parcel geometries, while workspace experiments achieved successful pickup at all 28 tested positions up to a 150 mm radial offset. End-to-end outdoor aerial collection achieved an 80% success rate (4/5 trials). These results demonstrate the feasibility of mechanically synchronized, winch-deployed grasping for landing-free autonomous aerial parcel collection.
Read the full abstract ▾
All authors Nimantha Adikaram, Mahen Abeyratne, Lakmina Chandrajith, A. H. T. E. De Silva, Isira Naotunna, Asanka Perera
Cryptography and Security
Safeguarding LLMs via Model-Agnostic Latent Safety Signals from Dark Knowledge
Wonjun Lee, Kyungsik Yang, Gaeun Ji et al. (+5 more)
LLMs have advanced rapidly, raising growing concerns about their safety. Recent work has proposed approaches to detect and defend against attacks including defenses at decoding stage that leverage models' hidden states. However, existing decoding-stage defenses suffer from two limitations. First, they introduce a trade-off between safety and over-refusal, where strengthening safety degrades the model's helpfulness on benign queries. Second, many of these methods rely on internal hidden states and are thus restricted to specific architectures, incurring substantial overhead and limited generalization across models. To address these limitations, we introduce LADE (Latent Safety Signals for Defense), which leverages latent safety signals extracted by contrasting harmful and benign queries from dark knowledge (i.e., information carried by the output probability distribution beyond its argmax) in the first-
token output probability distribution. Our key insight is that, beyond surface-level refusal
tokens, the dark knowledge in the first-
token distribution contains latent safety signals, defined as
tokens whose probabilities differ sharply between harmful and benign queries. We show that these signals consistently align across LLMs, forming a model-agnostic direction that emerges from safety
alignment. LADE consists of three components: (1) Extracting Latent Safety Signals from Dark Knowledge, which selects top-k safety-discriminative
tokens from the first-
token probability distribution; (2) Tokenizer Mapping, which maps these
tokens across different tokenizers to enable model-agnostic application; and (3) kNN-based Discrimination, which classifies queries via a k-Nearest Neighbors search over the mapped
tokens. Across diverse LLMs and
benchmarks, LADE is robust against a wide range of jailbreak attacks and lowers attack success rates while maintaining a competitive safety-utility trade-off.
Read the full abstract ▾
All authors Wonjun Lee, Kyungsik Yang, Gaeun Ji, Vaidehi Patil, Haon Park, Bumsub Ham, Mohit Bansal, Suhyun Kim
Cryptography and Security
BVI: Lightweight, Data-Centric Blockchain-Based Verification of Identity Claims
Harshith Pothapala, Gagandeep Singh, Samudi Amarasinghe et al. (+2 more)
A person who answers an unexpected call claiming to come from a bank has no way to check the claim. Australian text messaging has labelled a message as unverified when the sender identifier is not registered since 1 July 2026, but a voice call still arrives with nothing behind it, and voice cloning has removed the last cue recipients relied on. We present BVI, which answers one question for the recipient: did the calling party prove it holds a credential a registered organisation issued for this call? BVI keeps the organisational record, its authorised channels and its revocation state on a public ledger in a directly queryable form, and puts all decision logic in the handset, which performs eleven checks, pays no transaction fee and holds no full-chain state. We define ten attack classes plus the case in which revocation cannot be resolved, compare them in a 10-by-4 coverage
matrix spanning BBCA, the content digest, the detector and composed BVI, and run 27 automated scenario tests against isolated Hardhat fixture states of the BVI contract; all 27 pass. For media substitution and synthetic
speech, the contract tests validate commitment and gating properties rather than claiming detector accuracy. The mandatory per-session anchor costs about 91.5k gas and is invariant in call duration and hop count; each participating intermediary additionally contributes one optional attestation write. In a 20,000-session-per-
cell path simulation, at one quarter carrier participation BVI detects 20.6 per cent of randomly located in-path rewrites versus 0.1 per cent for the every-hop comparison, while identity and content evidence remain available even when path evidence is indeterminate. We also quantify the throughput ceiling of the anchoring design.
Read the full abstract ▾
All authors Harshith Pothapala, Gagandeep Singh, Samudi Amarasinghe, Priyanka Singh, Xue Li
Machine Learning
Targeted search shows that random-device testing underestimates worst-case error in a simulated wave-based neural operator
Samrendra Roy, Jason Yoo, Souvik Chakraborty et al. (+1 more)
Wave-based processors promise fast, energy-efficient
Fourier layers for neural operators. They are usually validated on randomly sampled devices, but using them requires knowing how large their error can become under fabrication and
alignment variation. In a stylised numerical case study, a hybrid
Fourier neural operator runs its four spectral layers on simulated coherent 4f processors with 32 toleranced knobs, whose half-widths are representative rather than calibrated. For 120 models (four tasks, six training methods, five seeds), we compared the worst of N random in-spec devices with a searched one. On a deterministic simulator with one frozen draw of the random static errors, the searched device's held-out error was 1.08-3.10 times the maximum over 200 Monte Carlo devices and 1.06-2.71 times that over 1000. With 20 fresh static draws, it still exceeded the maximum over 200 random devices in 116 of 120 models. Under uniform sampling, the probability of drawing such a device is at most 0.37% per model (two-sided 95% Clopper-Pearson), which says nothing about how large its error is. The gap persisted with uniform or Sobol' sampling at the search's budget, shared knobs, a second crosstalk model, box scales of 0.25-2 and a pixel-level device model. Models trained only with random static errors reached 3.7-39.9 times their nominal error on searched devices, and fine-tuning on random and gradient-searched devices gave the lowest searched error of the six in all 20 task-seed pairs. For two heat-exchanger quantities, a search targeted at each exceeded the worst of 1000 random devices in all 39 models, and hence the Wilks 95/95 limit (worst of 59). For the mean pressure of 11 models, no random device exceeded a 1% error threshold, but the searched device did. Random testing estimates how often errors exceed a threshold; worst-device search gives a lower bound on how large they can be.
Read the full abstract ▾
All authors Samrendra Roy, Jason Yoo, Souvik Chakraborty, Syed Bahauddin Alam
Robotics
Task-Space Imitation Guidance for Efficient Reinforcement Learning
Salar Asayesh, Hossein Darani, Todd Cao et al. (+2 more)
We introduce Task-Space Imitation Guidance for Efficient Reinforcement Learning (TIGER), a reward-construction and pretraining framework for sparse-reward tabletop robotic manipulation. TIGER treats an action-chunked imitation policy not as an executable controller or action prior, but as a local task-space progress
estimator: predicted action chunks are converted, using controller-aware action-to-motion mapping, into short-horizon end-effector references, and the RL
agent receives dense progress rewards toward these references while the sparse environment reward remains the dominant objective. During pretraining, TIGER uses imitation-guided look-ahead signals to relax conservative value penalties for actions predicted to make task-space progress, reducing off-
manifold exploration during early online RL. Across simulation and real-
robot experiments, TIGER improves early sample efficiency and reduces measured safety violations while matching or improving final success rates relative to prior RL and IL-RL baselines on the evaluated tasks.
Videos and an overview of this work are available at https://sanctuary.ai/news/tiger/.
Read the full abstract ▾
All authors Salar Asayesh, Hossein Darani, Todd Cao, Evgeny Andriash, Mani Ranjbar
Robotics
ReDex: Repairing Sim-to-Real Dexterous Policies by Finger-Level Compliant Interaction
Jinzhou Li, Hadi Tabatabaee, Kelin Yu et al. (+6 more)
Dexterous manipulation policies trained in simulation often fail to transfer to the real world because of errors in contact timing and force regulation. Yet these policies can retain useful multi-finger coordination for task progression. We propose ReDex, a framework for adapting a simulation-trained base policy to the real world by correcting local contact failures and incorporating tactile feedback. Starting from a proprioception-only base policy, ReDex allows a human operator to physically correct contact failures at selected fingers under compliant control during real-world rollouts, while the frozen base policy continues to control the remaining fingers. These rollouts combine base policy execution, human-corrected finger motion, and fingertip force observations. We reconstruct force-informed targets from these rollouts to train a standalone force-conditioned policy via behavior cloning. This design reduces human correction effort, enables learning of contact regulation from real-world interaction, and introduces force feedback into a proprioception-only policy without tactile simulation or complex full-hand teleoperation. We evaluate ReDex on two challenging, contact-rich dexterous manipulation tasks on real hardware. Compared with sim-to-real transferred base policies, ReDex increases Object Flipping success rate from 14% to 86% across two objects and average Screwdriver Rotation progress from 26.0% to 95.3% across three objects.
Read the full abstract ▾
All authors Jinzhou Li, Hadi Tabatabaee, Kelin Yu, Yuyin Sun, Cheng-Hao Kuo, Roberto Martín-Martín, Nima Fazeli, X. Alice Wu, Xianyi Cheng
Human-Computer Interaction
PAIR: Perceptual Affective Inference and Regulation in a Real-Time Multimodal Conversational Agent
Kexin Quan, Zijian Ding, Jiaye Yong et al. (+3 more)
Sustained emotional support requires generative
agents to connect momentary
emotion inference and regulation with continuity across encounters. We present PAIR (Perceptual Affective Inference and Regulation), a real-time
multimodal agent that reconstructs how an event is appraised into an emotional state. Appraisal scaffolds produce a valence-arousal-dominance estimate and select regulation guidance, delivered through conversation with coordinated
speech, color, and avatar cues. Rolling
memory carries context across sessions, and the scaffold re-runs after guidance. In a 14-day deployment with 19 participants, 1,093 sessions paired initial and post-guidance estimates with unanchored self-reports. Initial valence reached MAE 1.20 on the 9-point SAM scale (r=.68), dominance reached MAE 1.30, and arousal showed weak agreement even after coarsening. Self-reported emotional change varied with initial state, with the largest valence increases in sessions that began at negative valence. Perceived understanding was associated with greater valence increase and showed little correspondence with numerical prediction error. Over two weeks, helpfulness increased while input shortened; interviews traced personalization and companionship to relevant recall, context updates, and familiar dialogue. These findings connect inference accuracy to conversational and temporal patterns of support through per-event, first-person evaluation.
Read the full abstract ▾
All authors Kexin Quan, Zijian Ding, Jiaye Yong, Qinshi Zhang, Dong Wang, Jessie Chin
Machine Learning
Activation Denoising: A Robustness View on Parallel vs Sequential LLM Quantization
Yan Scholten, Rachel Lawrence, James Hensman et al. (+3 more)
Post-training quantization is a powerful tool for compressing large
language models. The most scalable methods quantize every layer in parallel, but quantization errors then compound through the residual stream, as no layer corrects for the errors of the layers before it. Sequential quantization accounts for this error compounding by re-calibrating each layer on the already-quantized outputs of its predecessors, yielding stronger results but at the cost of a serial schedule that becomes a bottleneck at scale. As a solution, we propose parallel quantization with activation denoising, which recovers much of the sequential benefit while keeping quantization fully parallel. Rather than re-calibrating layer-by-layer, we take a robustness perspective and model the upstream error as noise, regularizing to be robust to it through a preprocessing step followed by metric-weighted rounding. Applied at every layer, this regularization forms a depth-compounding smoothness penalty that dampens how strongly quantization errors amplify through the model. Unlike orthogonal rotations commonly used in quantization, which must preserve the model's function, we multiply the weights by a more general linear transformation. We find that the two are complementary and their effects compound. Empirically, our robustness regularization recovers a significant part of sequential quantization's benefit in a single parallel pass, at a fraction of its time. Overall, by treating compounding quantization errors as a robustness problem, we offer a principled foundation for more efficient and accurate LLM quantization at scale.
Read the full abstract ▾
All authors Yan Scholten, Rachel Lawrence, James Hensman, Stephan Günnemann, Alicia Curth, Riccardo Grazzi
Artificial Intelligence
Grounding What Shapes the Plan: Rethinking Groundedness for Physical Intelligence in Autonomous Driving
Minkyoung Cho, Zewei Zhou, Wenhao Ding et al. (+11 more)
Driving models increasingly ground
reasoning in causal relations, spatial structure, perceptual evidence, and predicted futures. These advances make
reasoning more faithful to the driving scene, but leave a fundamental question unresolved: what should groundedness mean when the model ultimately outputs an action? Correctly grounded
reasoning does not, by itself, ensure desirable driving outcomes. We introduce GroundAct, which starts from a simple premise: driving unfolds through physical entities and their interactions. Entities therefore become the unit of grounding; a lightweight reference
token keeps each selected entity's continuous state addressable through symbolic
reasoning; and only the referenced entities' interactions with the evolving proposal correct the plan. The result is an explicit path from what
reasoning grounds to what the plan does, which we call grounded planning. To assess its practical value, we evaluate GroundAct in both open- and closed-loop settings. GroundAct shows strong open-loop planning across normal, out-of-distribution, and safety-critical scenarios, with closed-loop results extending this evidence to driving in simulation.
Read the full abstract ▾
All authors Minkyoung Cho, Zewei Zhou, Wenhao Ding, Shuhan Tan, Boyi Li, Yuxiao Chen, Yan Wang, Zheng Lian, Min-Hung Chen, Chaowei Xiao, Zhuoqing Mao, Boris Ivanovic, Marco Pavone, Yulong Cao