Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

AcademicHub aggregates real-time literature from top journals and preprint platforms. Build your personal research radar and let large language models compile cross-disciplinary analysis briefings automatically.

01.
arXiv (CS.AI) 2026-06-15

When Good Verifiers Go Bad: Self-Improving VLMs Can Regress on New Tasks

Authors:

arXiv:2606.14629v1 Announce Type: cross Abstract: Verifier-driven self-DPO is a common recipe for self-improving production visual-language models. In this setup, a frozen verifier scores candidate generations, the top- and bottom-scoring candidates form a preference example, and DPO updates the learner. The deployment-time assumption is monotone: a stronger verifier should yield a stronger student. We show that this assumption can fail because verifier quality is highly task-specific. On a four-rung open-source verifier ladder across MathVista, MMMU, and BLINK, the same verifiers that are above-threshold and improve a Qwen-3-VL-2B student on MathVista become sub-threshold on MMMU, where their task-rubric accuracy drops to 8% to 23%. In this regime, every verifier we tested silently regresses the student, producing drops of 3.4 to 10.9 percentage points below the frozen baseline while the DPO training loss continues to decrease. The regression replicates on a second student, Qwen-2.5-VL-3B. Moreover, within the failure regime, damage is confidence-inverted: the more accurate-but-still-wrong verifier causes larger regression than a near-random verifier, suggesting that progress-gated replay amplifies confidently wrong preference pairs. We give a compact mechanistic explanation via a variance theorem for progress-gated replay and its direction-mismatch failure mode. The deployment message is operational rather than purely diagnostic: before running any verifier-driven loop, teams should measure target-task rubric accuracy, rank verifiers by target-task rubric quality rather than parameter count, and treat diminishing returns in above-threshold regimes as a verifier-side compute budget cap.

02.
arXiv (CS.CV) 2026-06-19

One-Shot Novel View and Pose Human Image Synthesis via 3D Prior Guided Diffusion Model

This paper addresses the challenge of one-shot novel view and pose human image synthesis. The existing methods transfer the reference human image to a target pose using a set of 2D pose keypoints or synthesize human images based on generalizable human NeRF which uses human model priors to extract point-wise features. However, pose transfer based methods can not handle complex human pose using ambiguous 2D pose as the condition, while generalizable human NeRFs may be inaccurate to recover occluded/invisiable human parts without extracted reliable features. To solve these problems, we propose a novel approach for novel view and pose synthesis from a singe human image via conditional denoising diffusion model. Our diffusion model divides the novel view and pose synthesis problem into a sequence of conditional denoising steps. Specifically, to generate humans with complex and arbitrary poses, we introduce 3D human priors, i.e., 3D normal map and color prompt, as geometry and color conditions into the generation process. By transferring the reference human into the target human with a series of diffusion steps, our diffusion model enables high-quality synthesis including the occluded/invisible parts. Further, we propose a self-reconstruction based customized refinement to enhance fine details when tested on novel persons.Experimental results on different public datasets demonstrate that our approach significantly outperforms previous methods and also shows better generalization ability across datasets. The code will be made publicly available at https://github.com/Yankeegsj/3DPGDM.

03.
arXiv (CS.CV) 2026-06-16

Tool-IQA: Augmenting Image Quality Assessment with Simple Tools

Vision-Language Models (VLMs) have been increasingly adopted for Image Quality Assessment (IQA). However, current methods typically employ a static one-shot scoring paradigm, despite the fact that humans assess image quality through dynamic visual inspection, e.g., selectively adjusting views to verify details and subtle artifacts. Specifically, relying solely on a single-pass observation introduces two primary limitations: first, perceiving the image only at a global scale restricts the assessment of finer local details; second, the original intensity distribution of the image may overwhelm the visibility, leading to insufficient inspection of image quality. To address these issues, we propose Tool-IQA, shifting the assessment mechanism from passive scoring to a tool-augmented workflow. In particular, we equip VLMs with simple yet effective view tools: a Magnifier to inspect local details, and a Gamma Corrector to uncover visibility and hidden artifacts. The assessment follows a structured pipeline that consists of an initial observation with rubric notes, a tool-augmented in-depth inspection, and a final quantification for calibrated quality score. Furthermore, to ensure efficient and purposeful tool callings, we introduce a batch-aware training strategy to reward tool interactions that can yield positive contributions rather than simply encouraging usage. Experiments on a variety of IQA benchmarks demonstrate that, with effective tool calling and calibrated assessment, our proposed Tool-IQA significantly outperforms existing state-of-the-art models, e.g., it achieves a PLCC of 0.854 on the challenging CLIVE dataset.

04.
arXiv (CS.LG) 2026-06-16

Identification and Inference for Algorithmic Frontiers with Selective Labels

arXiv:2606.14977v1 Announce Type: cross Abstract: This paper provides identification results to characterize a fairness-accuracy (FA) frontier, and statistical inference tools to test hypotheses and build a confidence set for the FA-frontier, when outcomes are observed only for selected individuals. When the selection process is unrestricted but loss is measured in specific ways, we provide a characterization of the sharp identification region of the FA-frontier. Under an assumption of unconfoundedness conditional on observables (and unrestricted loss functions), we obtain point identification and propose a debiased machine learning estimator, derive its asymptotic distribution, and show how this can be used to carry out inference for the FA-frontier. In work in progress, we extend the partial identification results to a broader class of loss functions.

05.
PLOS Medicine 2026-05-14

Antibody fine specificity correlates with protection from malaria for the RTS,S vaccine in young African children: A post hoc analysis of a phase IIb randomised controlled trial

Authors:

by Alessia Hysa, D. Herbert Opi, Joshua Waterhouse, Sandra Chishimba, Jessica L. Horton, Natalie Kingston, Hans J. Netter, David Wetzel, Michael Piontek, Gaoqian Feng, Jahit Sacarlal, Carlota Dobaño, Liriye Kurtovic, James G. Beeson Background The RTS,S/AS01 malaria vaccine was recently approved for implementation in children, but only provides modest and short-lived efficacy against malaria. RTS,S targets a portion of the Plasmodium falciparum (Pf) circumsporozoite protein (CSP), comprising the central NANP-repeat region and C-terminal domain. Mechanisms of immunity and correlates of protection for the RTS,S vaccine are not well defined, hindering progress towards generating highly effective CSP-based vaccines. Methods and findings We investigated epitope specificity and cross-reactivity of vaccine-induced antibodies to six peptides representing CSP epitopes in the N-terminal and central NANP-repeat region. We evaluated antibody reactivity in preclinical mouse vaccine studies, among CSP-specific monoclonal antibodies (mAbs), and in a large RTS,S phase IIb clinical trial in young children 1–4 years old (n = 735).The preclinical mouse vaccine studies and CSP-specific mAbs were used to initially evaluate IgG responses to the six peptides. Mice immunised with the central NANP-repeat region had IgG with cross-reactivity to an epitope in the N-terminal region. Additionally, we demonstrated that a single CSP-specific mAb could display cross-reactivity to several CSP epitopes. Through post hoc quantification and analysis of antibody responses in the RTS,S phase IIb clinical trial, we found that a subset of children generated IgG with specificity for a short NANP-repeat epitope (NANP2; amino acid sequence: NANPNANP) and cross-reactivity to an N-terminal epitope (J1; amino acid sequence: KQPADGNPDPNANPN). Notably, children with high IgG responses to NANP2 and J1 had a significantly reduced risk of clinical malaria, compared to children with low responses (IgG to NANP2 (aHR: 0.838 (95% CI [0.716, 0.981]; p = 0.028)) and J1 (aHR: 0.718 (95% CI [0.611, 0.844]; p 

06.
arXiv (CS.LG) 2026-06-16

DemoDiffusion: One-Shot Human Imitation using pre-trained Diffusion Policy

arXiv:2506.20668v3 Announce Type: replace-cross Abstract: We propose DemoDiffusion, a simple method for enabling robots to perform manipulation tasks by imitating a single human demonstration, without requiring task-specific training or paired human-robot data. Our approach is based on two insights. First, the hand motion in a human demonstration provides a useful prior for the robot's end-effector trajectory, which we can convert into a rough open-loop robot motion trajectory via kinematic retargeting. Second, while this retargeted motion captures the overall structure of the task, it may not align well with plausible robot actions in-context. To address this, we leverage a pre-trained generalist diffusion policy to modify the trajectory, ensuring it both follows the human motion and remains within the distribution of plausible robot actions. Unlike approaches based on online reinforcement learning or paired human-robot data, our method enables robust adaptation to new tasks and scenes with minimal effort. In real-world experiments across 8 diverse manipulation tasks, DemoDiffusion achieves 83.8\% average success rate, compared to 13.8\% for the pre-trained policy and 52.5\% for kinematic retargeting, succeeding even on tasks where the pre-trained generalist policy fails entirely. Project page: https://demodiffusion.github.io/

07.
arXiv (CS.AI) 2026-06-11

"That's AI Slop, You Bot!" Studying Accusations, Evidence, and Credibility in Online Discourse Towards LLM-Generated Comments

arXiv:2606.12073v1 Announce Type: cross Abstract: Generative AI has made fluent prose cheap to produce, breaking the old promise to readers that good writing meant real thinking. How have readers responded, and what can this tell us about changing anti-AI attitudes? We analyzed 25 million comments from Hacker News and Reddit (2023-2026), combining LLM judgment on 7,500 sampled accusations of AI use, sentiment trajectories, speech-act coding of 300 confirmed accusations of AI use, and a matched-control test of accused versus non-accused parent comments. We found that the pejorative-label share of accusations rose more than tenfold on both platforms while a placebo vocabulary of pre-2022 inauthenticity terms (shill, astroturf) did not. This shift reflected a fast-growing trend of branding any suspicious or seemingly inauthentic prose as "AI slop". The slop frame now constitutes 94 percent of pejorative mentions, with the dominant comments shifting in tone from mockery toward gatekeeping and structural protest. The key surprise comes from a matched-control test which found that prose features that statistically distinguish AI from human text do not predict which human text gets accused as AI. The new accusations work as social gatekeeping of perceived authenticity without actually screening for AI. This research extends signaling theory by showing that substitute signals used socially can grow even when inaccurate if the underlying detection problem cannot be solved at the non-expert level. It shows that AI's effects on writing from the reader side are distinct from those on the production (writer) side. Detection technology cannot resolve this dynamic because the social function of accusations is increasingly to perform social gatekeeping and in-group signaling as opposed to identifying AI-generated writing.

08.
arXiv (CS.CV) 2026-06-12

Radar-Guided Polynomial Fitting for Metric Depth Estimation

We propose POLAR, a novel radar-guided depth estimation method that introduces polynomial fitting to efficiently transform scaleless depth predictions from pretrained monocular depth estimation (MDE) models into metric depth maps. Unlike existing approaches that rely on complex architectures or expensive sensors, our method is grounded in a fundamental insight: although MDE models often infer reasonable local depth structure within each object or local region, they may misalign these regions relative to one another, making a linear scale and shift (affine) transformation insufficient given three or more of these regions. To address this limitation, we use polynomial coefficients predicted from cheap, ubiquitous radar data to adaptively adjust predictions non-uniformly across depth ranges. In this way, POLAR generalizes beyond affine transformations and is able to correct such misalignments by introducing inflection points. Importantly, our polynomial fitting framework preserves structural consistency through a novel training objective that enforces local monotonicity via first-derivative regularization. POLAR achieves state-of-the-art performance across three datasets, outperforming existing methods by an average of 24.9% in MAE and 33.2% in RMSE, while also achieving state-of-the-art efficiency in terms of latency and computational cost.

09.
arXiv (CS.LG) 2026-06-18

On the Residual Scaling of Looped Transformers: Stability and Transferability

arXiv:2606.18524v1 Announce Type: new Abstract: Looped (weight-tied) Transformers apply a shared residual block $N$ times ($h \leftarrow h + \varepsilon\,f(h)$, same $f$ at each step), increasing effective depth without adding parameters. Prior depth-scaling analyses prescribe $\varepsilon = 1/\!\sqrt{L}$ for depth-$L$ residual networks. We show that this is insufficient for looped architectures: weight sharing makes residual updates correlated across iterations, requiring the stronger scaling $\varepsilon = 1/N$. For multi-layer blocks ($L$ unique layers looped $N$ times), we derive a factored parameterization $\varepsilon = \lambda/(N\!\sqrt{L})$ that separates the two sources of growth: $1/N$ controls the within-layer loop correlation, and $1/\!\sqrt{L}$ controls the across-layer variance. A key consequence is that the optimal learning rate depends only on the number of unique layers $L$, not on the loop count $N$, enabling direct hyperparameter transfer from small to large $N$ without retuning. Experiments on looped Transformers confirm that $1/N$ scaling improves trainability and yields better loss than $1/\!\sqrt{N}$ scaling across loop counts.

10.
arXiv (CS.CL) 2026-06-25

OPERA: Aligning Open-Ended Reasoning via Objective Perplexity-based Reinforcement Learning

Reinforcement Learning (RL) has enabled LLMs to excel in objective reasoning tasks such as mathematics and code generation. However, applying RL to open-ended tasks, such as creative writing, remains challenging because LLM-as-a-judge reward models often exhibit stylistic biases and positional inconsistencies, leading to unstable supervision. To address this, we propose OPERA (Objective Perplexity-based Reflective Alignment), which replaces unreliable external judges with intrinsic rewards derived from perplexity dynamics. Specifically, we derive an intrinsic reward signal from perplexity dynamics, quantifying uncertainty reduction at critical reflective states. During the cold-start phase, we introduce a data synthesis method that leverages carefully designed guiding words to generate diverse reasoning traces, along with perplexity-prioritized rollouts that utilize internal log-probabilities to identify logically consistent reasoning branches. This pipeline yields a large-scale dataset comprising 20,000 high-quality reasoning trajectories. Empirical evaluations consistently demonstrate the scalability and efficacy of our approach in alignment for open-ended tasks. Implementing OPERA on Qwen3-8B establishes a new state-of-the-art among open-source models, achieving parity with or surpassing proprietary models like Gemini2.5 and MiniMax-M2.5 in some open-ended tasks. The code is available at https://github.com/pangpang-xuan/OPERA.

11.
arXiv (quant-ph) 2026-06-15

Quantum Entanglement of Bethe States

arXiv:2606.14140v1 Announce Type: cross Abstract: We investigate the quantum entanglement of Bethe states across a family of integrable spin chains, including the XXX$_{\frac{1}{2}}$ model, its higher-spin generalizations (XXX$_s$), and the non-compact $SL(2,\mathbb{R})$ chain. For on-shell eigenstates, we perform a comprehensive scan of the bipartite entanglement entropy across the entire spectrum of finite chains with periodic boundary conditions, and identify the Bethe solutions that minimize and maximize the entanglement. These extremal solutions follow systematic, spin-dependent patterns in the Bethe quantum numbers. In the XXX$_{\frac{1}{2}}$ spin chain, for the antiferromagnetic chain, the state with minimal entropy always coincides with the lowest-energy state (the ground state) within a given fixed-magnon sector. For the higher-spin XXX$_s$ model, however, the lowest-entropy state is not always identical to the ground state, and can even be the state of highest energy. By contrast, the Bethe roots that maximize entropy exhibit considerably more intricate structure. Our analysis further reveals how special Bethe root configurations, such as singular and strange solutions, affect entanglement, and it uncovers characteristic entanglement features in the non-compact $SL(2,\mathbb{R})$ chain that are absent from compact spin chains. For off-shell Bethe states, we develop an optimization algorithm that extremizes the entanglement entropy over rapidity distributions, enabling us to explore the maximum entanglement achievable by a Bethe state without imposing the Bethe ansatz equations.

12.
arXiv (CS.CV) 2026-06-15

RATS! Patches Talk Through Registers: Emergent Parts in Register Attention Transformers

When humans see a bird, they recognize far more than just "bird" – they see a head, wings, and talons, a structured assembly of reusable parts that can be identified across every bird they have ever seen. We ask whether a self-supervised visual model can discover the same compositional structure on its own. To this end, we propose RATS (Register Attention Transformers), which decomposes the classification token into N learnable register tokens that route patch information through an L->N->N->L bottleneck via a three-step compress-communicate-broadcast attention. The N registers are partitioned across the H attention heads, so that registers assigned to different heads do not interact with each other. Without auxiliary losses or part annotations, each register spontaneously specializes into a proto-semantic region whose emerging structure resembles object parts. RATS surpasses all baselines by +12 mIoU on average across five segmentation benchmarks, with consistent gains on ADE20K (+1.11 mIoU) and COCO (+0.2 AP^m). Its register dictionary further exhibits part-level consistency and semantic proximity across related categories. Our results suggest that RATS may provide a useful architectural prior for structured and interpretable visual representation learning.

13.
arXiv (CS.AI) 2026-06-12

Iterating Toward Better Search: A Two-Agent Simulation Framework for Evaluating Agentic Search Architectures in E-Commerce

arXiv:2606.12924v1 Announce Type: new Abstract: We present a modular two-agent simulation framework for evaluating conversational shopping assistant architectures. An independent buyer agent, configured with personas, missions, and patience levels, is paired with an interchangeable responder that integrates with a real e-commerce search API. Holding the buyer constant across experiments enables controlled comparison of responder designs on identical scenarios. Using 2011 conversations across 14 persona buckets, we establish four empirical findings. First, rolling-window memory outperforms intent-extraction memory on all quality metrics while being 35% faster per query. Second, illustrating rapid evidence-driven iteration, a systematic failure analysis of a responder version enables targeted fixes that reduce failure and near-failure rates by 62% across the full dataset. Third, swapping the responder LLM backbone from Gemini~2.5 to Llama~3.3~70B costs 0.16–0.45 points despite identical architecture. Finally, we document systematic philosophical disagreement between frontier LLM judges: Gemini rewards process correctness while Claude demands concrete outcomes, despite using the same evaluation prompt.

14.
arXiv (CS.LG) 2026-06-24

A Robust Model-Based Approach for Continuous-Time Policy Evaluation with Unknown Lévy Process Dynamics

arXiv:2504.01482v3 Announce Type: replace-cross Abstract: This paper develops a model-based framework for continuous-time policy evaluation (CTPE) in reinforcement learning, incorporating both Brownian and Lévy noise to model stochastic dynamics influenced by rare and extreme events. Our approach formulates the policy evaluation problem as solving a partial integro-differential equation (PIDE) for the value function with unknown coefficients. A key challenge in this setting is accurately recovering the unknown coefficients in the stochastic dynamics, particularly when driven by Lévy processes with heavy tail effects. To address this, we propose a robust numerical approach that effectively handles both unbiased and censored trajectory datasets. This method combines maximum likelihood estimation with an iterative tail correction mechanism, improving the stability and accuracy of coefficient recovery. Additionally, we establish a theoretical bound for the policy evaluation error based on coefficient recovery error. Through numerical experiments, including a real-data BTC price experiment, we demonstrate the effectiveness and robustness of our method in recovering heavy-tailed Lévy dynamics and verify the theoretical error analysis in policy evaluation.

15.
arXiv (CS.AI) 2026-06-24

Assessing Distribution Shift in Human Activity Recognition for Domain Generalization

arXiv:2606.24781v1 Announce Type: new Abstract: While the field of Human Activity Recognition (HAR) continues to draw interest from researchers and advance in important ways, some key challenges remain. One of the most difficult aspects of building HAR models that show good performance in real-world settings is dealing with data diversity from device and sensor heterogeneity, and contextual changes that are intrinsic to real-world applications. While data diversity in HAR has been well-acknowledged in the literature, there remains a gap in understanding the effect of various types of distribution shifts on HAR models and the domain generalization problem that arises. Towards that end, this paper systematically evaluates 4 different types of distribution shifts, including variations in device type, sensor placement, sampling rate, and user behavior. Quantifying their effects, we illustrate that diversity shifts predominantly define all types of shifts, indicating the existence of unique features that are not shared across different domains. We then introduce a uniform HAR-based distribution shift benchmarks and conduct a comprehensive evaluation of up to 28 domain generalization methods. Our analysis exposes the limitations of current domain generalization algorithms in achieving model generalizability, marginally outperforming the empirical risk minimization baseline. This work represents the first systematic exploration of domain generalization and adaptation concerning specific distribution shifts in sensor-based HAR, offering an open-source benchmark platform and datasets to spur further research.

16.
arXiv (CS.LG) 2026-06-16

Privacy from Symmetry: Orthogonally Equivariant Transformers for LLM Inference

arXiv:2606.16461v1 Announce Type: new Abstract: Running large language models locally is often impractical, pushing inference on sensitive text to third-party providers. Split inference partially mitigates this by keeping tokens on the client and sending only hidden representations, but these representations can still be recovered via nearest-neighbor search against the public embedding table. We propose an orthogonal obfuscation procedure in which the client multiplies embeddings by a secret orthogonal matrix before transmission. To enable correct inference under arbitrary rotations, we introduce ConjFormer, a transformer variant that is exactly $\mathrm{O}(d)$-equivariant via a lightweight normalization change (scalar RMSNorm) together with blockwise orthogonal conjugation of all linear weights. As a result, the server performs the full forward pass entirely in the rotated basis and never observes unrotated hidden states. Experiments on GPT-2 and Llama 3.2 1B models fine-tuned on PubMed show that orthogonal obfuscation eliminates direct cosine nearest-neighbor inversion and reduces token recovery from over 35% top-10 to at most 1.3%, while increasing perplexity by only 0.4% after fine-tuning. These results indicate that enforcing symmetry at the architectural level can provide a practical defense for privacy-preserving LLM inference without noise injection or heavy cryptographic machinery.

17.
arXiv (CS.LG) 2026-06-12

How Reliable are Fairness Audits with Unreliable Data?

arXiv:2506.23033v3 Announce Type: replace Abstract: Fairness audits are a key component of responsible machine-learning deployment. Yet, audit-recommendation reliability under incomplete protected-label access is still poorly understood. In this work, we focused on protected-label missingness in fairness mitigation audits. We introduced a seed-calibrated stress test to separate missingness effects from seed-to-seed movement already present under complete labels. Across ACS/Folktables tasks, missingness settings that retain some protected labels usually do not move selected mitigation methods beyond a complete-label seed-to-seed baseline. At $0%$ protected-label access, candidates collapse to an empirical-risk-minimization baseline and deterministic tie-breaking rather than revealing a broad missingness effect. We also found that threshold optimization can turn fairness gains on a single protected axis into intersectional harm above a seed baseline, and this threshold-optimizer finding persists under random-forest validation. Overall, our results highlight that protected-label missingness should be reported with seed-null calibration, candidate-set context, and intersectional consequences before it is treated as evidence of audit fragility.

18.
arXiv (CS.CV) 2026-06-25

$S^{2}$-FracMix: Label-Preserving Self-Saliency Mixup Augmentation

Data augmentation is known to improve generalization of deep visual models. Recent methods favor mixup strategies that generate interpolated samples to improve model performance. However, these techniques not only incur significant computational overhead, they also lead to semantic disruption of augmentation data due to cross-sample mixing. We first propose Self-Saliency ($S^2$) Mixup, which constructs challenging yet label-consistent samples by extracting multi-scale salient patches and reinserting them into non-salient regions of the same image. This promotes scale-invariant feature learning while avoiding cross-sample interference. To further enhance model robustness, we introduce FracMix, a mixing scheme that injects self-similarity patterns into salient regions using adaptive ratios. Collectively, our unified framework, $S^{2}$-FracMix, enables simultaneous learning from fractal and non-fractal structures within a single image, yielding a targeted and structurally coherent augmentation strategy. We theoretically analyze the advantage of our technique, and empirically establish its superiority over the existing methods by achieving state-of-the-art performance in extensive evaluation with seven benchmarks across classification (coarse and fine-grained), robustness, calibration, object detection, and transfer learning tasks. Project page is available at \href{https://fracmix-data-augmentation.github.io/}{fracmix-data-augmentation.github.io}

19.
arXiv (CS.LG) 2026-06-19

On the Variance of Temporal Difference Learning and its Reduction Using Control Variates

arXiv:2606.20357v1 Announce Type: new Abstract: We analyze the variance of temporal difference (TD) learning using the phased setting with tabular representation, and show that one of the mechanisms behind its ability to reduce variance is by effectively aggregating over a larger number of independent trajectories. Based on this insight, we demonstrate that (1) the variance of TD is asymptotically bounded from above by Monte Carlo (MC) estimators, and (2) shorter horizon updates incurs less variance for a fixed number of samples. Beyond TD, we show that Direct Advantage Estimation (DAE), a method for estimating the advantage function, can be seen as a type of regression-adjusted control variate, which achieves a tighter bound on the variance compared to TD in the large-sample limit. Finally, we numerically illustrate the behaviors of these estimators with carefully designed environments.

20.
arXiv (CS.AI) 2026-06-18

Veriphi: Attack-Guided Neural Network Verification with Dataset-Dependent Training Methods

arXiv:2606.18454v1 Announce Type: cross Abstract: We present Veriphi, a GPU-accelerated neural network verification system that combines fast adversarial attacks with formal bound certification using alpha,beta-CROWN methods. Through systematic experiments on MNIST and CIFAR-10 using three training methodologies (standard, adversarial, certified), we demonstrate that training method effectiveness is fundamentally dataset-dependent. Interval Bound Propagation (IBP) achieves 78% certified accuracy on simple MNIST (784 dimensions) but provides negligible certification performance on the more complex CIFAR-10 dataset, where PGD adversarial training dominates with 94% certification at small perturbations. We achieve 5x verification speedup through attack-guided falsification and scale our approach to production-size models (105.8M parameters) for real-world aerospace logistics optimization. Our results challenge the assumption that certified training universally outperforms adversarial training, showing context matters critically for verification strategy selection.

21.
arXiv (CS.LG) 2026-06-25

Learning with Monotone Adversarial Corruptions

arXiv:2601.02193v2 Announce Type: replace Abstract: We study the extent to which standard machine learning algorithms rely on exchangeability and independence of data by introducing a monotone adversarial corruption model. In this model, an adversary, upon looking at a "clean" i.i.d. dataset, inserts additional "corrupted" points of their choice into the dataset. These added points are constrained to be monotone corruptions, in that they get labeled according to the ground-truth target function. Perhaps surprisingly, we demonstrate that in this setting, all known optimal learning algorithms for binary classification can be made to achieve suboptimal expected error on a new independent test point drawn from the same distribution as the clean dataset. On the other hand, we show that uniform convergence-based algorithms do not degrade in their guarantees. Our results showcase how optimal learning algorithms break down in the face of seemingly helpful monotone corruptions, exposing their overreliance on exchangeability.

22.
arXiv (CS.CV) 2026-06-12

Possibilistic Predictive Uncertainty for Deep Learning

Deep neural networks achieve impressive results across diverse applications, yet their overconfidence on unseen inputs necessitates reliable epistemic uncertainty modeling. Existing methods for uncertainty modeling face a fundamental dilemma: Bayesian approaches provide principled estimates but remain computationally prohibitive, while efficient second-order predictors lack rigorous connections between their specific objectives and epistemic uncertainty quantification. To resolve this dilemma, we introduce Dirichlet-approximated possibilistic posterior predictions (DAPPr), a principled framework grounded in possibility theory. We define a possibilistic posterior over parameters, project it to the prediction space via supremum operators, and approximate the projected posterior using learnable Dirichlet possibility functions. This projection-and-approximation strategy yields a simple training objective with closed-form solutions. Despite its simplicity, extensive experiments across diverse benchmarks show that DAPPr achieves competitive or superior uncertainty quantification performance over state-of-the-art second-order predictors while maintaining both principled derivation and computational efficiency. Code is available at https://github.com/MaxwellYaoNi/DAPPr.

23.
arXiv (CS.LG) 2026-06-19

Sparsity, Superposition, and Forgetting: A Mechanistic Study of Representation Retention in Continual Learning

arXiv:2606.20431v1 Announce Type: new Abstract: Continual learning (CL) systems often forget previously acquired knowledge, yet the mechanisms driving forgetting remain hard to isolate in practice because real datasets entangle many factors. We present a controlled, toy-world framework that makes these mechanisms observable and testable. Using a synthetic generator-separator pipeline, we define ground-truth latent features, build tasks with tunable sparsity and overlap, and introduce measurable quantities for representation strength and superposition (directional overlap among features). We then study retention dynamics-the temporal change of representation strength by fitting sparse dynamical relations (via SINDy) between retention, superposition, and exposure history. A complementary task-level analysis based on effective rank characterizes how representational capacity is allocated across tasks. Our controlled experiments yield three takeaways. (1) Superposition tends to increase over time with transient dips at task boundaries, suggesting boundary-specific interference rather than steady drift. (2) Higher feature sparsity induces more superposition yet does not inevitably cause forgetting; when representations remain strong, forgetting can be reduced despite overlap. (3) Task-level effective rank grows with sparsity, indicating broader capacity usage under sparse regimes. Together, these results nuance the common intuition that more superposition leads to more forgetting by showing that overlap interacts with representation strength and capacity allocation. Our toy analysis provides falsifiable hypotheses and diagnostic tools for CL.

24.
arXiv (CS.LG) 2026-06-15

Behavioral Audit of Machine Unlearning Has a Privacy Cost

arXiv:2606.14518v1 Announce Type: new Abstract: The removal of learned data from Machine Learning models through Machine Unlearning (MU) has been widely studied; however, there has yet to be an agreed-upon scheme for auditing MU. Existing work has shown that a dishonest model owner can falsify evidence to avoid executing MU, while curious auditors (and adversaries) can infer the privacy-sensitive properties of the model and its training data even with limited access. Yet auditing of MU under mutual distrust between the model owner and the auditor remains unexplored. We provide an information-theoretic proof for this scenario: for convex ML models, a generic audit scheme that relies solely on querying the model for behavioral signals cannot identify insufficiently unlearned models without revealing membership information of the retained set. Therefore, auditing MU under the assumption of a dishonest model owner and an honest-but-curious auditor faces an inherent privacy-audit tradeoff. Our empirical results on convex models strongly supports this result, while further experiments demonstrate that this privacy-audit tension persists in non-convex models. Our results call for a more careful consideration of the privacy-audit tension under a realistic auditor threat model, and serve as a foundation for more scrutiny of designs of privacy-preserving audit schemes for the MU pipeline. We also release our code implementation at https://github.com/LiouTang/Behavioral-Unlearn-Audit.

25.
arXiv (CS.CL) 2026-06-18

Efficient Financial Language Understanding via Distillation with Synthetic Data

Large instruction-following models are powerful but costly to deploy, particularly in finance, where labelled data are limited by confidentiality and expert annotation cost. We present an efficient framework for financial sentiment analysis through distillation with synthetic data, transferring knowledge from a large instruction-tuned teacher to compact student models. The framework is designed for low-resource conditions, where a small set of real examples are collected and labelled by hand. The framework then clusters the examples and uses the clusters to select seeds for generating synthetic examples via structured few-shot prompting. Experiments show that clustering-based seed selection yields more representative synthetic data than random sampling, enabling compact models to achieve strong performance with minimal supervision. Notably, on a more complex and noisy text domain, the compact model trained on the complete synthetic-seed corpus even outperforms the teacher model, while remaining competitive on formal text. The framework provides a practical route toward resource-efficient domain adaptation in financial NLP with minimal human labelling effort.