Robotics
Beyond Policy Alignment: Closing the Planning-Learning Loop for Robot Control with Learned World Models
Kowndinya Boyalakuntla, Yuhan Liu and Abdeslam Boularias
Planning with learned world models combines online trajectory optimization with learned value and policy functions for high-dimensional control. Because the planner determines the experience used for learning, while the learned critic and actor in turn score and propose future plans, planning and learning form a closed feedback loop. TD-MPC is a prominent instance of this design. Recent policy-constrained variants strengthen one part of the loop by aligning the learned policy with planner behavior. We introduce PL-MPC (Planning-Learning MPC), which additionally modifies critic supervision and planner terminal-value estimation. Hybrid multi-step TD targets expose critic updates to more realized rewards before bootstrapping; disagreement-aware terminal estimates reduce the influence of uncertain critic values during MPPI planning; and return-weighted actor
distillation emphasizes planner-executed actions from high-return episodes. The world-model architecture and MPPI optimizer are otherwise unchanged. On HumanoidBench, the largest gains occur on balance-hard, where Total Average Return (TAR) increases from $98\pm18$ to $387\pm255$, and hurdle, from $199\pm13$ to $466\pm200$; performance across the broader
benchmark remains task dependent, and PL-MPC remains competitive on DMControl. Controlled ablations show different component interactions across the two tasks. We further demonstrate zero-shot sim-to-real transfer on wrench-nut
alignment with a 7-DoF KUKA IIWA14, obtaining higher observed success than TD-M(PC)^2 on the training object size and two unseen sizes. Code and data will be available at: https://pl-mpc-
humanoid.github.io.
Read the full abstract ▾
Human-Computer Interaction
Engagement-Led Segmentation of Gamified Participation Data in a Large-Scale Remote Internship: A Mixed-Methods Study
Sakshi Sharma, Pavani Ayinampudi, Aditya B.M.V. et al. (+5 more)
Gamified points are widely used to represent learner participation, but cumulative point totals provide limited information about how participation changes over time. This limitation matters in continuously enrolling programmes, where learners have different opportunities to accumulate points. This study examines how gamified participation data can be interpreted as indicators of behavioural engagement in a large-scale, continuously enrolling remote internship. Using an anonymised operational
dataset, 3,607 distinct started learners were first classified by participation status, separating dormant learners from those with observable participation. The 1,871 active learners were then grouped using K-means clustering on normalised longitudinal and opportunity-aware participation features. Four participation patterns emerged: Thriving Core, Tapering, Fast Starters, and Occasional Participants. Their trajectories differed in both level and direction. A perception survey of 597 respondents provided complementary learner-reported evidence, integrated with the behavioural strand through a joint display by segment. Perceptions differed across segments, while individual-level correlations between perceptions and behaviour were small, and learners with low recorded participation could report positive perceptions of the points system alongside external barriers. Segments assigned from the first four weeks were then checked against eleven weeks of later platform records: 94% of the Thriving Core attended at least ten further sessions and 22% completed the internship, against under 7% and under 1% in the low-engagement segments. The findings suggest that gamified participation data are more informative when interpreted as longitudinal, opportunity-aware behavioural indicators rather than cumulative scores alone.
Read the full abstract ▾
All authors Sakshi Sharma, Pavani Ayinampudi, Aditya B.M.V., Jinal Gupta, Prakash Hegade, Rohit Sharma, Meenakshi V, S.R.S. Iyengar
Machine Learning
Validity-Preserving Hierarchical RL for Joint Routing and Switch Placement in EDA
Dorian Gailhard, Ugo Lecerf, Enzo Tartaglione et al. (+2 more)
Routing and switch placement are fundamental combinatorial optimization problems in chip design, requiring the joint optimization of routing
topology and physical placement under strict structural,
geometric and logical constraints. Existing approaches typically rely on carefully engineered heuristics that incorporate strong problem-specific biases to navigate the enormous space of possible designs. In this work, we introduce a hierarchical reinforcement learning framework for joint routing and switch placement at the level of logical communication routes. Starting from a minimal routing graph, our method progressively constructs increasingly expressive solutions through three coupled operations: switch expansion, switch placement, and route refinement. These operations preserve routing validity by construction, restricting exploration to feasible configurations where every communicating initiator-target pair has one assigned loop-free route. We explore the induced solution space using Gumbel Monte Carlo Tree Search, showing that neural-guided search substantially improves solution quality over non-learning optimization methods. Furthermore, pretraining across floorplans provides a strong initialization for fine-tuning on unseen instances.
Read the full abstract ▾
All authors Dorian Gailhard, Ugo Lecerf, Enzo Tartaglione, Donatello Conte, Jhony H. Giraldo
Computer Vision
Yongjian Zhang, Longguang Wang, Zhuo Song et al. (+3 more)
Scaling has become a primary driver of progress in
language and
vision foundation models, yet its role in precise correspondence matching remains underexplored. In this work, we present Flow Any Scene
Transformer (FAST), a scalable correspondence model driven by two key insights. First, we reveal that the query-key projections inside single-view
vision foundation models encode a coarse yet reusable prior for cross-view matching. Second, reusing these pretrained projections in cross-
attention form yields a highly effective initialization for a ViT-based matcher built from a single-view encoder. Guided by these insights, we build FAST upon a vanilla single-view foundation model, utilizing a zero-parameter rewiring strategy to convert selected self-
attention layers into cross-
attention for cross-view interaction. This design allows ViT-based matchers to scale with advances in single-view foundation models, bypassing the need for a dedicated pair-centric pretraining stage. To fully unlock the scaling potential of this formulation, we assemble a 6-million-pair training corpus for general-purpose dense 2D displacement estimation across diverse co-visible image pairs. Extensive experiments demonstrate that FAST achieves state-of-the-art performance across a wide range of
benchmarks, while scaling favorably with both backbone size and training data.
Read the full abstract ▾
All authors Yongjian Zhang, Longguang Wang, Zhuo Song, Zhiheng Fu, Liang Lin, Yulan Guo
Information Theory · Cross-listed from quant-ph
A depolarizing choir sings in Gaussian harmony
Rabsan Galib Ahmed, Sujeet Bhalerao, Sungjai Lee et al. (+4 more)
We study the noise threshold for positive
quantum capacity for the
qubit depolarizing channel. We explore analytically the action of the
qubit depolarizing channel on the symmetric subspaces of the input
qubits, in the limit of asymptotically many uses of the channel. We observe the emergence of a bosonic Gaussian channel. Furthermore, the codes previously developed for the depolarizing channel can be translated to codes for the emergent Gaussian channel, and it is easier to further optimize these codes for the simpler emergent Gaussian channel. Translating these codes back to the depolarizing channel leads to extremely good input states for the coherent information of the depolarizing channel producing new lower bounds on the noise threshold for positive capacity. In addition to improved lower bounds on the threshold, this newly found link between depolarizing noise and Gaussian channels offers a novel perspective contributing to our understanding of these symmetric codes.
Read the full abstract ▾
All authors Rabsan Galib Ahmed, Sujeet Bhalerao, Sungjai Lee, Felix Leditzky, Debbie Leung, Luke Schaeffer, Graeme Smith
Machine Learning
The Nixtlaverse: An Open-Source Ecosystem for Forecasting
Olivier Sprangers, Max Mergenthaler Canseco, Marco Peixeiro et al. (+8 more)
Large
forecasting applications often combine statistical, machine-learning, and neural models. These families solve the same problem but differ in fitted state, training procedures, and how they parallelize work.
Forecasting software must therefore either hide these differences behind a single
estimator interface, or keep the families in separate packages, forcing users to rewrite data preparation and evaluation for every package. We present the Nixtlaverse, an ecosystem of open-source Python libraries for time series
forecasting, as a case study of a third design: all libraries share the same long-format panel data and keyed forecast outputs, while every model family keeps its own specialized implementation. We demonstrate this design through three use cases on the public M5 competition data. First, we evaluate statistical, machine-learning, and neural models, and an external engine from a separate ecosystem, in a single rolling-origin evaluation with per-series and hierarchy-weighted metrics. Second, we profile runtime and peak
memory from 100 to 30,490 series and locate each family's bottleneck: statistical fitting scales approximately linearly in the number of series, feature construction dominates machine-learning
memory, and neural training time is nearly independent of panel size under a fixed training budget. Third, we reconcile the forecasts of multiple engines, including the external one, over all 42,840 series of the M5 hierarchy, with sparse reconciliation where dense implementations exhausted
memory. These use cases establish the costs, boundaries, and utility of shared data and output contracts. The Nixtlaverse has seen substantial public distribution, scholarly reuse, and adoption through other
forecasting frameworks, and is released under permissive open-source licenses with public
datasets, reproducible examples, and verifiable
benchmark artifacts.
Read the full abstract ▾
All authors Olivier Sprangers, Max Mergenthaler Canseco, Marco Peixeiro, Saul Caballero Ramirez, Mariana Menchero García, Jing-Qiang Goh, Han Wang, Nikhil Gupta, Rogelio Melo, Senbong Gee, Cristian Challu
Computation and Language
LatentHarness: Learning Latent Actions for Memory and Reasoning via Counterfactual Policy Distillation
Xiaoqiang Wang, Suyuchen Wang and Bang Liu
Long-context
reasoning faces two complementary bottlenecks: retaining evidence across long inputs and sustaining computation across many
reasoning steps. Existing approaches largely address them separately, with external
memory extending access to distant evidence and latent
reasoning compressing multi-step computation. We introduce LatentHarness, which unifies
memory access and latent
reasoning as sequential latent action selection. At each internal step, the model chooses THINK for further computation, RECALL from a fast-weight
memory of input evidence and intermediate
reasoning states, or EXIT to emit the next
token. We train this policy with counterfactual policy
distillation, which branches every action for one step and scores its effect on the emitted
token. These gains teach the policy when
memory is more useful than further
reasoning, while gradients through counterfactual recall teach which intermediate states should be retained in
memory for future use. Across six general and long-context
reasoning benchmarks, LatentHarness at 1.4B improves on the strongest baselines by 2.8% and 10.0% relative, respectively, and runs 5.9x faster than the strongest long-context baseline.
Read the full abstract ▾
Machine Learning
Stable Transformers for Graph Generation
Luca Miglior, Alessio Gravina and Davide Bacciu
Graph generative models increasingly rely on Graph
Transformers (GT) to capture complex dependencies among nodes and edges. While deeper architectures should provide greater expressive capacity and a broader receptive field, their effectiveness can decline with depth: repeated self-
attention progressively contracts node representations, impeding information flow and gradient propagation. We analyse this phenomenon from a dynamical systems perspective, focusing on how the denoiser's spectral dynamics affect graph generation. We show that standard GT denoisers become increasingly dissipative as depth grows, leading to vanishing gradients and representation collapse. To isolate the effect of these dynamics, we construct a permutation-equivariant GT with inherently stable, non-dissipative transport. We also introduce a damping mechanism that continuously interpolates between non-dissipative and increasingly contractive regimes, enabling a direct assessment of how dissipation influences generation. Experiments on synthetic and molecular graph generation
benchmarks show that the gap between these regimes widens with depth: non-dissipative dynamics preserve representation diversity and gradient flow, sustaining strong generative performance, whereas greater contraction progressively impairs it. These findings identify the denoiser's dynamical regime as a key design factor for deep graph generative models.
Read the full abstract ▾
Machine Learning
CNCGEN: A Dataset and Framework for Machining Process Planning and Toolpath Generation from B-rep Models
Xiaolei Zhou, Boyi Lin, Yuchao Feng et al. (+1 more)
Learning to generate machining process plans and toolpaths from B-rep CAD requires coupling discrete operation decisions with continuous tool motion as the workpiece evolves. Correctly predicting an operation sequence does not by itself ensure correct material removal, because each toolpath acts on the stock left by preceding cuts. We formulate this problem around persistent manufacturing objects: object identity determines the target of an operation, while the evolving stock state conditions the generation of its toolpath. Based on this formulation, we propose CNCGEN, a
dataset and learning framework for three-axis machining. CNCGEN-
Dataset contains approximately 50k geometrically verified synthetic machining flows and 800 held-out real CNC records. Each flow aligns B-rep geometry with object-referenced operations, parameterized toolpaths, intermediate stock states, and verification outcomes, enabling supervision of the correspondence between planning decisions and their
geometric effects. CNCGEN generates operations and toolpaths for selected objects step by step, updating a compact machining state to guide subsequent predictions. During training, a learned surrogate verifier provides material-removal feedback that links local predictions to their
geometric consequences. Experiments on synthetic and held-out real CNC records show that CNCGEN improves the resulting workpiece geometry and reduces residual material and overcut compared with adapted CNC generation baselines.
Read the full abstract ▾
All authors Xiaolei Zhou, Boyi Lin, Yuchao Feng, Jianwei Zheng
Machine Learning
A library for differentiable signal processing and machine learning on the sphere
Thorsten Kurth, Max Rietmann, Mauro Bisson et al. (+6 more)
The two-dimensional sphere embedded in three-dimensional Euclidean space S2, plays a central role in a variety of scientific and engineering domains, including geophysics, planetary science, geodesy, atmospheric physics,
quantum chemistry, cosmology, and virtual reality, among many others. As machine learning increasingly permeates these fields, the demand grows for robust tools that process and model functions on the sphere, while respecting the inherent topological and
symmetry properties of the domain. We present torch-harmonics, a comprehensive library that offers efficient, differentiable implementations of advanced signal processing and machine learning (ML) methods for spherical data. These include the spherical harmonic transform (SHT), the spherical analogue of the
Fourier transform, vector spherical harmonics, discrete-continuous and spectral convolutions, as well as both global and neighborhood spherical
attention mechanisms. Beyond traditional representations, torch-harmonics provides the building blocks for state-of-the-art spherical ML architectures such as spherical
transformers in order to enable scalable, rotationally-aware learning and inference in modern scientific and engineering applications.
Read the full abstract ▾
All authors Thorsten Kurth, Max Rietmann, Mauro Bisson, Andrea Paris, Alberto Carpentieri, Jean Kossaifi, Anima Anandkumar, Christian Hundt, Boris Bonev
Artificial Intelligence
OverForge: Reasoning Through Strategies and Tactics Helps Cooperative Lifelong Adaptation
Oana Madalina Fron, Ojas Shirekar and Chirag Raman
Cooperative
language-model
agents must coordinate over long horizons and adapt to changing environments and to partners with unfamiliar conventions, yet existing
agents map observations to actions without separating persistent coordination strategies from their tactical execution. We introduce OverForge, a training-free hierarchical architecture that separates strategic
reasoning over roles and divisions of labour from tactical
reasoning over actions within each
agent's private, partner-conditioned world model. A metacognitive Prefrontal Cortex Module couples the two levels by forming strategy-action branches, imagining their consequences with a forward model, and committing when confident. In OvercookedV2, OverForge delivers 7 soups in a connected kitchen versus 3 for each flat LLM baseline, retains agreed roles, and adopts roles proposed by unfamiliar partners. Ablations and a fixed-strategy probe show that persistent strategies guide tactical adaptation while each
reasoning level contributes to coordination.
Memory restarts show that cross-episode partner knowledge supports task performance and partner prediction, linking the hierarchy to continual adaptation.
Read the full abstract ▾
Discrete Mathematics
Solving polynomial equations over partial discrete dynamical systems
Maximilien Gadouleau, Sara Riva, Marius Rolland et al. (+1 more)
The analysis of observable phenomena (for instance, in biology or physics) allows the detection of dynamical behaviours. If the conditions are ideal and the number of observations is sufficient, we can represent these phenomena by a dynamical system, also called a functional digraph, that is to say a graph where each node has out-degree exactly one. Up to isomorphism, these dynamical systems, equipped with disjoint union as addition and direct product as multiplication, form a commutative semiring. Several previous studies on this semiring have aimed to establish algebraic properties (primality, injectivity) or complexity results (division, factorisation). However, no work has yet been conducted on graphs derived from imperfect observations that result in missing transitions or nodes, in other words, in cases where each node in the graph has an out-degree of at most one. In this case, we say that the system is partial. In this paper, we show that partial dynamical systems, up to isomorphism and equipped with the same addition and multiplication, still form a commutative semiring. We then characterise the
prime elements of this semiring, which differ from those of the semiring of dynamical systems. Finally, we highlight two properties shared by both semirings. First, injective univariate
polynomials admit the same characterisation in both. Second, division can be computed in
polynomial time for partial dynamical systems if and only if it can be for dynamical systems.
Read the full abstract ▾
All authors Maximilien Gadouleau, Sara Riva, Marius Rolland, Marie-Emile Voge
Computer Vision
Let the Carrier Carry the Attack: Preserving the Subject in Adversarial Image Generation
Linfeng Jiang, Steven McDonagh, Yuhang Chen et al. (+3 more)
Strong unrestricted
adversarial attacks can distort the primary object of an image, hereafter referred to as the subject. To preserve subject integrity without compromising attack magnitude, we introduce the carrier: a secondary visual element that provides an auxiliary region to facilitate the attack under global classifier guidance. We demonstrate three key findings: 1. A carrier mitigates subject distortion by absorbing a larger share of globally normalized attack updates. 2. A carrier improves cross-model transferability, governed by the strength of target-related features that balance semantic separation and transfer performance. 3. Successful targeted attacks retain the personalized subject as the primary content perceived by humans while successfully misleading the classifier. Our results demonstrate that a visually secondary carrier offers an auxiliary spatial pathway for
adversarial changes, enabling strong and transferable attacks while improving subject preservation.
Read the full abstract ▾
All authors Linfeng Jiang, Steven McDonagh, Yuhang Chen, Xingyu Zhao, Siddartha Khastgir, Andi Zhang
Sound
Synthetic Speech Attribution via Prototypical Networks
Viola Negroni, Paolo Bestagini and Stefano Tubaro
Synthetic
speech attribution aims to identify the generative system responsible for a
speech signal, but current approaches typically rely on black-box neural networks that provide limited insight into their decisions. This work investigates prototype-based networks as an interpretable alternative, where predictions are grounded in comparisons with representative training examples. We adapt ProtoPNet to spectrogram-based
speech representations and evaluate the proposed framework on the MLAAD
dataset under closed-set, cross-lingual, and open-set conditions. Experiments show that prototype-based
reasoning achieves competitive or improved attribution performance compared with the baseline while enabling example-based explanations. These results highlight that interpretability and performance can be jointly achieved in synthetic
speech attribution through prototype-based modeling.
Read the full abstract ▾
Systems and Control
Transient Triggering Grid-Forming Synchronization Control Under Voltage and Frequency Dips
Dewan Mahnaaz Mahmud, Vinu Thomas, Bogdan Marinescu et al. (+1 more)
Grid-forming (GFM) inverters are gaining
attention as a promising alternative for conventional synchronous generators in the modern power systems. Unlike conventional synchronous generators, GFM inverters have limited overcurrent capability that makes them vulnerable during large disturbances. During disturbances i.e., voltage and frequency dips, GFM inverters are pushed into current-limited operation to protect the
semiconductor switches. This causes the internal angle of the GFM inverters to accelerate and lose synchronism with the rest of the grid. To address this limitation, this article proposes a transient triggering grid-forming (TTGFM) synchronization control to enhance the synchronization stability performance under voltage and frequency dips. This method uses two feedback signals; terminal voltage and the difference between unsaturated and saturated power to adjust the internal angle which is generated by power synchronization loop (PSL) of the GFM inverters. These two signals manipulates the internal reference angle generation that act as a virtual braking mechanism. This mechanism limits the angle acceleration during voltage and frequency dips without requiring an extra supervisory signal or parameters to tune. The proposed method is benchmarked against two state-of-the-art synchronization enhancement schemes and validated through high-fidelity electromagnetic transient (EMT) simulations with a grid dynamic equivalent (GDE) model in MATLAB/Simulink\textsuperscript{\textregistered}. An analytical framework is developed to derive the synchronization instability mechanism and the critical limits of the stability margins. Generalization of the GDE model further shows that the TTGFM control is not restricted to a single configuration but is applicable to any standard
benchmark system.
Read the full abstract ▾
All authors Dewan Mahnaaz Mahmud, Vinu Thomas, Bogdan Marinescu, Mickaël Hilairet
Computational Geometry
Parameterized Hardness of Mixed 2-Sided Orthant Depth
Michelle Döring and Georg Tennigkeit
We consider the maximum-depth problem for mixed 2-sided orthants in R^d: each region imposes one lower bound and one upper bound on distinct coordinates, and the task is to find a point contained in as many regions as possible. We show that the corresponding decision problem is W[1]-hard when parameterized by the dimension. Our reduction from MultiColoredClique uses two coordinates per color class and only polynomially many orthants.
Read the full abstract ▾
Artificial Intelligence
Trust Is Not a Score: Runtime Assurance Contracts for High-Risk AI Agents
Serhii Zabolotnii
Benchmarks, audits, and
agent protocols describe performance, permissions, and repair, but not how observed evidence should change an
agent's authority during a consequential task. We propose a Runtime Assurance Contract (RAC), a policy-level formal schema binding autonomy boundaries, component eligibility, evidence state, transition policy, human-review capacity, and non-compensatory gates. Under RAC, soft metrics may inform routing, whereas a failed or unknown mandatory gate forces retry, switch, escalation, deferral, or stop; aggregate performance cannot authorize action. We define the contract, an evidence record, a permission rule, and five invariants. A deterministic failure-injection study in agentic coding evaluates 174 cases, each a distinct mechanism, under a gate conjunction, a score-only rule, and a restricted protocol baseline. At the published example weights and threshold, the score rule admits 56 of 63 block-required injections and all 24 review-required injections; tuned in hindsight, it matches the conjunction case for case. For positive weights, a positive threshold, binary risk signals, and each signal firing alone on some case, the two rules agree on every case if and only if the threshold does not exceed the smallest weight. On 124 further cases written blind by a model of another vendor, the conjunction itself admits 35 of 79 injections, and so does the tuned score rule: a missing detector is invisible to every decision rule. Eighteen hand-authored traces check version-pinned evidence and review transitions. In a prospective synthetic holdout of 24 episodes, RAC and a separately implemented full stateful baseline both match the labels of two blinded LLM judges. These studies test mechanisms on synthetic cases; they establish neither deployed safety nor cross-domain effectiveness.
Read the full abstract ▾
Distributed Computing
Exploring Adaptive Byzantine Quorum Systems to Improve Latency in the WAN
Linus Gnan, Rüdiger Kapitza and Christian Berger
Quorum systems enforce strict consistency in Byzantine fault-tolerant (BFT) state machine replication: Before a value is decided, a subset of replicas (called quorum) must exchange votes for the value. In wide-area networks, the size and composition of a quorum determines the speed at which replicas can make progress and thus impacts the latency perceived by clients. A variety of quorum constructions has been proposed, e.g., threshold, weighted, grid and others, but the literature offers little in the way of an apples-to-apples comparison. Each construction typically lives inside a different protocol implementation, so the influence of quorum design on client latency cannot be easily separated from other implementation effects, especially the use of optimization techniques. In this paper, we close that gap by extending BFT-SMaRt with a modular quorum-system abstraction that allows a variety of quorum constructions to live inside the same framework and interact with the same set of optimization techniques. Via our framework, we explore the impact of different quorum system constructions on client-observed latency in wide-area BFT replication using network simulation.
Read the full abstract ▾
Artificial Intelligence
ArchitectureIQ: On the Measure of Training Intuition
Zirui Ren, Shaoyang Guo, Chencheng Tang et al. (+9 more)
Top researchers have good intuition, but do
language models have as good intuition about model training as top AI researchers? To measure model intuition of LLMs and humans, we introduce the ArchitectureIQ
benchmark. Each question presents a synthetic
dataset and several training recipes, and the test-taker is asked to predict the recipe yielding the best test metric. Overall, we find that LLMs' model intuition is good but has four limitations: (1) The intuition is imperfect, or even sub-human in some cases. Frontier models achieve around 76% accuracy (random choice 33%) vs best human researcher (66.0%), yet remain far from perfect. For architecture-only questions, best human achieves 65% while GPT-6 Astra only has 38%. (2) The intuition is empirical, not structured, supported by the fact that more CoT compute does not lead to substantial improvement. Unlike math, we still lack a "Science of AI"
language that enables structured
reasoning on AI. (3) The intuition is not maximally condensed, and can be further compressed into a knoledge base. Our constructed knowledge base with only 20 items yields large gains for weak models: GPT-4o equipped with the accumulated knowledge almost matches the performance of Claude Opus 5. (4) The intuition is insensitive to
dataset properties, but the best model should in general depend on data properties. This suggests that data is the real "dark matter" in AI – LLMs (so do human researchers) understand too little about data, even less than model architectures.
Read the full abstract ▾
All authors Zirui Ren, Shaoyang Guo, Chencheng Tang, Jinxin Wang, Chengyu Xiong, Shanbin Yu, Peihang Li, Yidi Wu, Bangzhe Huang, Qingyu Qu, Leqian Yang, Ziming Liu
Computation and Language
Drift Inspector: Exploring and Measuring Scientific Drift with Atomic Contribution Claims
Vsevolod Karimov, Stepan Ostarkov, Anastasia Poroshina et al. (+2 more)
Scientific abstracts mix contributions with background, motivation, and meta-
language, so tools that read them as-is cannot separate what a field produces from what it discusses. We present Drift Inspector, an open-source system for measuring and exploring how a research field changes over time at the level of Atomic Contribution Claims (ACCs): decontextualized, contribution-bearing propositions an LLM extracts from each abstract before analysis. The system clusters these claims across years into an interactive map where every trend traces back to the claims and papers behind it. Applied to six years of EMNLP, it shows the field shifting away from classic NLP tasks toward LLM-era capabilities such as
reasoning and multimodality – a movement that keyword or whole-abstract counts blur. The released data extend beyond EMNLP: the same pipeline has processed the full ACL Anthology (346k claims, 80k abstracts, 423 venues). Extraction is human-validated and clustering checked against an external manually constructed taxonomy.
Read the full abstract ▾
All authors Vsevolod Karimov, Stepan Ostarkov, Anastasia Poroshina, Anatoly Frolov, Alexander Panchenko
Computer Vision
BTC3D: Blended Tile Conditioning for Detail-Enhancing Image-to-3D Generation
Junyu Li, Qiuyu Chen, Pengcheng Wang et al. (+5 more)
Recent
diffusion-based pipelines have achieved promising progress in image-to-3D synthesis. However, generating high-fidelity details remains challenging, especially when the input image contains rich details. Existing approaches often rely on globally encoded conditioning features, which compress spatial information and limit the model to reproduce fine-grained details. This common design often leads to a phenomenon we term detail attenuation. Moreover, improving image-to-3D synthesis quality typically requires retraining or fine-tuning large
diffusion models, which can be computationally expensive and impractical for complex 3D pipelines. In this work, we present Blended Tile Conditioning for image-to-3D generation (BTC3D), a training-free inference time framework that enhances fine-grained detail preservation in image-to-3D
diffusion pipelines. To alleviate detail attenuation, we first examine the image feature additivity in image-to-3D models. Based on this property, we introduce a blended tile
embedding that extracts local conditioning signals from split image regional patches, allowing the
diffusion model to better preserve fine-grained visual details. To integrate the global and local conditioning guidance stably, we propose a dynamic conditioning schedule that gradually increases the influence of tile-level conditioning during later low-noise stages of
diffusion. Our proposed method BTC3D operates entirely at inference time and can be seamlessly integrated into existing image-to-3D
diffusion pipelines. Experimental results demonstrate that the proposed approach significantly improves texture quality and visual fidelity of the base model while maintaining global structural consistency in a training-free manner.
Read the full abstract ▾
All authors Junyu Li, Qiuyu Chen, Pengcheng Wang, Shiqi Yang, Alexandra Gomez-Villa, Joost van de Weijer, Ruilin Li, Kai Wang
Information Theory · Cross-listed from quant-ph
The Intrinsic Cost of Quantum Syndrome Extraction
Chia-Tung Chu
How much time is fundamentally required to extract the syndrome of a stabilizer code on hardware with a given connectivity graph? Previous bounds answer through a chosen list of checks (Delfosse, Beverland, and Tremblay, 2021) or through the global code parameters (Baspin, Fawzi, and Shayeghi, 2023). Here, we show that neither determines the interaction that a particular hardware cut requires. We therefore define the intrinsic cut cost, $\chi$, directly from the stabilizer group, independent of the generating set used to describe the code. It is computable from the check
matrix alone, and equals the mutual information across the cut in the maximally mixed code state. We show that every exact extraction
circuit, however adaptive, needs at least $\chi/2$ two-
qubit operations across the cut, and alternatively with free local processing exactly $\chi$ shared Bell pairs. Guided by this intrinsic cut cost, we construct annular surface codes with constant-depth local extraction that give a counterexample to Corollary 1 of Delfosse, Beverland, and Tremblay. Furthermore, for a fixed
quantum Tanner family, $\chi$ grows at least linearly with the size of every sufficiently small region, giving a depth guarantee at each corresponding bottleneck in every layout. This strengthens the global perspective of Baspin, Fawzi, and Shayeghi and, together with our matching square-grid construction, yields a tight space–depth tradeoff. Our results connect the interaction required by syndrome extraction to hardware communication capacity and thus provide a way to assess unavoidable
circuit costs before committing to a particular layout.
Read the full abstract ▾
Computer Vision
When Masking Helps or Hurts Robustness in Compressed CLIP: A Pre-Deployment Diagnostic
Muhammad Zawish and Steven Davy
This paper demonstrate that whether masking-based
token pruning helps or hurts worst-group robustness can be predicted before deployment, without labels or fine-tuning. A systematic study of semantic masking across 8 spurious-correlation
benchmarks shows its effect on worst-group accuracy is highly unstable: it improves accuracy by up to 82.5% relative on some
datasets and degrades it by up to 100% on others. We trace this instability to spurious inversion: background patches receive higher CLIP text-similarity than the true object when the spurious attribute is background-separable, inverting the assumption every text- and
attention-guided
pruning method relies on. We introduce the Spurious Inversion Metric (SIM), a label-free, pre-deployment diagnostic whose sign predicts this effect with statistical significance (binomial $p=0.035$) across all 8
datasets, and remains dependable across 6 CLIP architectures with a clean foreground/background split. Naive masking is itself a major source of risk: it causes the largest average-accuracy loss of any method we evaluate, and its own per-image
segmentation step is a significant runtime bottleneck. To address this, we design a batched, synchronization-free GPU
segmentation routine that cuts this overhead from 3.5$\times$ to 1.75$\times$ baseline. Gating deployment by SIM's sign recovers masking's benefits while avoiding its worst failures, matching or exceeding a strong
pruning baseline on 7 of 8
datasets.
Read the full abstract ▾
Systems and Control
STELLA: A 16nm Spatio-Temporal Elastic Low-Latency CGRA for Multi-Stage Pipelined Applications
Jun Yin, Chao Fang, Ryan Antonio et al. (+4 more)
Emerging non-
matrix ML kernels, such as LayerNorm, GeLu, FFT or circular convolutions, demand low-latency, energy-efficient spatial accelerators beyond MatMul-centric arrays. STELLA presents a spatio-temporal elastic 16 nm coarse-grained reconfigurable array (CGRA) with a rapid configuration path, per-PE hardware loop control, and a low-latency, deeply pipelined elastic fabric with spatio-temporal data reuse. STELLA reaches up to 110 GOPS/mm2 at 850 MHz, and improves effective kernel throughput by 4.84-7.14x over baseline CGRAs.
Read the full abstract ▾
All authors Jun Yin, Chao Fang, Ryan Antonio, Xiaoling Yi, Yunhao Deng, Fanchen Kong, Marian Verhelst
Artificial Intelligence
A helps B while B hurts A: directed transfer in instruction-tuning mixture
Nima H. Siboni and Vahid Rostami
Adapting a
language model to a specialized corpus means choosing which instruction-tuning tasks to train on under a fixed budget, and testing one choice costs a fine-tuning run. Common heuristics add more source tasks or pick sources similar to the target. The first assumes transfer is never negative; the second, that it is symmetric. We show that both assumptions fail: task $A$ can help task $B$ while $B$ hurts $A$, so helpfulness is a signed property of ordered source–target pairs. We introduce the transfer map, a signed estimate of how much each source helps or hurts each held-out target. We fit the map in hundreds of fine-tuning runs on Qwen3 and Mistral models from 0.6B to 32B parameters, with all sources drawn from one corpus and no training examples from the target. The map predicts a held-out target's accuracy on unseen mixtures: recorded before those runs, its predictions have less than half the error of a mixture-agnostic baseline. The map is specific to its target and corpus but transfers across model scale: a mixture selected in advance at one size beats training on all source tasks at every other size we tested. Transfer is thus a property of the data. The map selects the tasks that help and drops the one that interferes: accuracy on the
reasoning targets (causal explanation, multi-hop questions and methodological critique) rises by up to 14 percentage points over training on all source tasks.
Read the full abstract ▾
Artificial Intelligence
Values as Style: Disentangling Values from Semantics with One-Way Mixing for Low-Damage LLM Steering
Jiale Dai, Hongcan Deng, Liuxian Ma et al. (+2 more)
Value steering should change an LLM's normative priorities while preserving the scenario, facts, and task constraints underlying its answer. Conventional activation edits often change both. We introduce an editable semantic-value interface on frozen residual states, with a one-way semantic-to-value pathway that grounds value recognition in context. Stop-gradient blocks feedback through this pathway; swap consistency, topic de-confounding, and decorrelation encourage selective codes. At inference, editing the value code produces a residual delta while holding the semantic code fixed. On two instruction-tuned backbones, this interface improves semantic preservation and reduces benign refusals at comparable value
alignment. A matched mixing-by-gating ablation separates representation learning from selective edit activation, and dimension-matched probes establish improved code selectivity. Against validation-selected prompting on LLaMA-3.1-8B, the method achieves comparable
alignment (0.750 vs. 0.748), higher BERTScore (0.938 vs. 0.923), and fewer contradictions (5.1% vs. 7.6%). Human ratings and cross-taxonomy controls provide complementary evidence for low-damage value steering.
Read the full abstract ▾
All authors Jiale Dai, Hongcan Deng, Liuxian Ma, Xiaoke Niu, Guojie Song
Information Retrieval
When the Label Ignores the Request: Auditing Policy-Selected Targets in Synthetic Conversational Music Recommendation
Sanjeev Suresh
Synthetic dialogues generated by LLM pipelines now serve as complete conversational-recommendation
benchmarks: an LLM listener talks to an LLM recommender, and the track logged next in the conversation becomes the official label for each turn. These policy-selected labels make large-scale evaluation reproducible, but they are proxies for what the simulated user asked. We audit the one place where label and request are directly comparable: turns where the user asks for an exact song by name. In the RecSys Challenge 2026 TalkPlay
benchmark, using visible dialogue and catalog metadata alone, we find that the official label contradicts the user's exact-song request in half of the audited development turns. This matters beyond one
benchmark: naming the desired item is the dominant intent in real music search, where deployed systems avoid substituting an alternative for an exactly named item, on the premise that it costs satisfaction. A small training-time supplement closes most of the gap: adding catalog-resolved request-satisfying targets to a small fraction of training turns yields a 53.3% relative gain in nDCG@20 on the 43 conflict turns while leaving the official metric intact, verified against a matched control that detects the same requests but trains only on official labels.
Read the full abstract ▾
Machine Learning
GFD-OPD: Guidance-Folded On-Policy Distillation of Diffusion Models Across Scales
Zhenxing Zhang, Jiayan Teng, Wenxu Wu et al. (+6 more)
On-policy
distillation (OPD) has demonstrated two important capabilities in
language models: compressing large teachers into smaller students and merging expert models into a single model. Existing
diffusion OPD, however, mostly focus on the latter, with teachers and students sharing the same backbone and scale. We investigate large-to-small
diffusion opd from large teachers to a small student and find that the standard recipe fails. To find the underlying cause, we propose Fixed-State KL, an effective and fair way to measure the distribution gap between student and teacher during OPD training for
diffusion models. We are the first to clarify why large-to-small OPD is challenging for
diffusion models: a smaller student struggles to perfectly match the distribution of a larger teacher, while classifier-free guidance can accumulate and amplify the distributional discrepancies between the student's conditional and unconditional branches and those of the teacher. To solve this problem, we propose GFD-OPD, a simple yet effective method that reduces the student-teacher gap while avoiding the error amplification of the CFG composition. Across numerous experiments, GFD outperforms previous baselines in both training efficiency and final performance, achieving state-of-the-art results on all
benchmarks.
Read the full abstract ▾
All authors Zhenxing Zhang, Jiayan Teng, Wenxu Wu, Zhuoyi Yang, Jiazheng Xu, Wendi Zheng, Jie Tang, Dan Guo, Meng Wang
Computer Vision
ShieldCLIP: Selective Safety Alignment for Harmful Content Mitigation in Multimodal Foundation Models
Tobia Poppi, Silvia Cappelletti, Samuele Poppi et al. (+4 more)
Multimodal encoders such as CLIP underlie many downstream systems, but their web-scale training data embed harmful associations that safety
alignment must suppress without unnecessarily changing benign representations. Because ethical and practical constraints prevent collecting real unsafe content at scale, existing
datasets pair safe real samples with generated counterparts, but label every generated sample unsafe, even when one modality is individually safe. To address this, we introduce ShieldCLIP, the first framework to condition safety
alignment on the observed safety state of each modality rather than the origin of a sample, preserving safe content while redirecting only what is unsafe. We also introduce ViSUv2, a 195k-quadruplet
dataset with independent per-modality safety labels across 578 concepts and 28 categories. Using these labels, ShieldCLIP defines a four-way conditional objective beyond pair-level supervision: safe content is anchored, unsafe modalities are redirected to their safe counterparts, mixed pairs update only the unsafe branch, and coherence is enforced when both are unsafe. We evaluate ShieldCLIP on cross-modal
retrieval, text-to-image generation with Stable
Diffusion v1.4 and SDXL, and image-to-text generation with LLaVA. Across these settings, ShieldCLIP consistently reduces harmful outputs over prior safety-aligned encoders and strong mitigation baselines, while preserving the utility of the original
embedding space. Extensive ablation studies further show that both modality-specific supervision and the selective
alignment objective contribute to these gains. Source code, trained models, and ViSUv2 (under a controlled-access protocol) will be made publicly available at https://aimagelab.github.io/ShieldCLIP/.
Read the full abstract ▾
All authors Tobia Poppi, Silvia Cappelletti, Samuele Poppi, Marcella Cornia, Lorenzo Baraldi, Diego Garcia-Olano, Rita Cucchiara
Computation and Language
Better Supervision Is Nearby: Neighborhood On-Policy Self-Distillation
Xincheng Wei, Yifan Ding, Yoshua Li et al. (+6 more)
On-policy self-
distillation (OPSD) trains mathematical
reasoning models using a privileged teacher that sees a reference solution and supervises student-sampled prefixes. Standard OPSD uses one fixed parameter setting at every state, but nearby settings may offer additional supervision. We find that local parameter perturbations reveal complementary reference-aligned corrections under the same reference context. Different experts supply these corrections at different reference positions. Their pool covers more such positions than the unperturbed privileged teacher. We introduce Neighborhood OPSD (N-OPSD) to turn these corrections into supervision at student-visited states. Offline, greedy selection builds a compact pool of frozen experts by rewarding filtered reference-
token gains beyond the pool's current best at each position. The highest-peak expert need not provide the best training target. Online routing therefore separates the anchor direction from its level of support. MaxPeak selects the anchor
token, and
quantile selection chooses among experts whose top
token matches it. The student learns from the chosen expert's full next-
token distribution through the clipped forward-KL objective inherited from OPSD. We evaluate on AIME 2024, AIME 2025, and HMMT February 2025. Across three independent runs per method, Neighborhood OPSD improves the three-
benchmark Average@12 over OPSD by 2.75, 1.67, and 1.94 points on Qwen3-1.7B, 4B, and 8B, respectively. Student-prefix continuations support using the pool beyond the reference trajectories used for selection. Matched ablations support filtered reference-
token gains as a selection criterion. Accounting for overlap within the pool and routing by state further improve student accuracy. Inference uses only the distilled student.
Read the full abstract ▾
All authors Xincheng Wei, Yifan Ding, Yoshua Li, Yuquan Lu, Ziheng Li, Yi Lu, Dongsheng Ma, Rongxiang Weng, Xunliang Cai