Academic Intelligence · Curated Daily

探索全球前沿学术脉络

AcademicHub 汇聚顶级期刊与预印本平台的实时文献。定制您的专属科研雷达,利用大语言模型自动生成交叉领域文献分析简报。

01.
arXiv (CS.CV) 2026-06-17

Colab NAS: Obtaining lightweight task-specific convolutional neural networks following Occam's razor

The current trend of applying transfer learning from convolutional neural networks (CNNs) trained on large datasets can be an overkill when the target application is a custom and delimited problem, with enough data to train a network from scratch. On the other hand, the training of custom and lighter CNNs requires expertise, in the from-scratch case, and or high-end resources, as in the case of hardware-aware neural architecture search (HW NAS), limiting access to the technology by non-habitual NN developers. For this reason, we present ColabNAS, an affordable HW NAS technique for producing lightweight task-specific CNNs. Its novel derivative-free search strategy, inspired by Occam's razor, allows to obtain state-of-the-art results on the Visual Wake Word dataset, a standard TinyML benchmark, in just 3.1 GPU hours using free online GPU services such as Google Colaboratory and Kaggle Kernel.

02.
arXiv (CS.AI) 2026-06-12

Reasoning as Pattern Matching: Shared Mechanisms in Human and LLM Everyday Reasoning

arXiv:2606.13607v1 Announce Type: new Abstract: When large language models (LLMs) fail to generalize or make haphazard errors in reasoning, it is often taken as evidence that LLMs are not truly reasoning, but rather performing a kind of pattern matching. The implication is that people's behavior does not exhibit the same types of failures because human reasoning uses principled and abstract world models. We evaluate human participants and 25 LLMs on their ability to engage in common-sense reasoning about a variety of everyday situations and observe similar patterns of errors in both people and models. We then identify the set of attention heads driving LLM responses and find that these heads implement a form of pattern-matching. These attention heads allow us to predict seemingly inexplicable reasoning errors in people caused by ostensibly irrelevant prompt details. Taken together, our results suggest that everyday causal reasoning in people and LLMs is more consistent with a form of pattern-matching than with abstract world models.

03.
arXiv (CS.LG) 2026-06-12

DeepJEB++: Foundation Model-Driven Large-Scale 3D Engineering Dataset via 2D Latent Space Augmentation

arXiv:2606.12994v1 Announce Type: new Abstract: Data-driven engineering design is constrained by the lack of large-scale 3D datasets that pair geometry with physics-based performance labels. In particular, existing 3D data augmentation techniques have limitations in preserving subtle and diverse geometric variations, and it remains difficult to automate the subsequent simulation-labeling process, where boundary conditions vary depending on the generated geometry. We present DeepJEB++, a foundation-model-driven data-augmentation framework that expands a small seed set of jet engine brackets into a large, simulation-labeled 3D dataset under constrained resources. Our key idea is to augment in the data-rich 2D latent space, then transfer to 3D. In Stage 1, we fine-tune a pretrained 2D latent diffusion model on multi-view renders and synthesize novel views by latent interpolation, retaining manufacturable designs through a vision-language-model (VLM) quality filter. In Stage 2, the validated images are lifted to 3D meshes by a domain-adapted generative foundation model. In Stage 3, an automated pipeline recognizes the load and bolt interfaces on each mesh and assigns finite-element labels – mass, stress, and displacement – without manual intervention. We assess augmentation quality along three intrinsic axes: manufacturability, label fidelity against the SimJEB ground truth, and distributional consistency. Starting from fewer than 400 seed designs, DeepJEB++ yields 15,360 simulation-labeled 3D brackets – a 40x expansion – using a single GPU per stage. The dataset will be made publicly available to support reproducible engineering-AI research.

04.
arXiv (quant-ph) 2026-06-17

The Standard Model, The Exceptional Jordan Algebra, and Triality

作者:

arXiv:2006.16265v2 Announce Type: replace-cross Abstract: Jordan, Wigner and von Neumann classified the possible algebras of quantum mechanical observables, and found they fell into 4 "ordinary" families, plus one remarkable outlier: the exceptional Jordan algebra. We point out an intriguing relationship between the complexification of this algebra and the standard model of particle physics, its minimal left-right-symmetric $SU(3)\times SU(2)_{L}\times SU(2)_{R}\times U(1)$ extension, and $Spin(10)$ unification. This suggests a geometric interpretation, where a single generation of standard model fermions is described by the tangent space $(\mathbb{C}\otimes\mathbb{O})^{2}$ of the complex octonionic projective plane, and the existence of three generations is related to $SO(8)$ triality.

05.
arXiv (CS.CV) 2026-06-15

MUSE: Agentic 3D Scene Authoring via Memory-Grounded Incremental Requirement Satisfaction

Text-driven 3D scene generation is a promising technique for digital content creation, embodied AI simulation, and interactive design, yet practical workflows often require refining, extending, or correcting existing scenes while preserving non-target content. Existing methods can produce realistic and structurally plausible scenes, but they generally lack editability with requirement-level state tracking, so part-level failures often lead to full-scene regeneration or manual intervention. To tackle this challenge, we formulate controllable 3D scene authoring as incremental requirement satisfaction, unifying construction and editing. In this paper, we present MUSE, a memory-grounded multi-agent framework in which an Architect compiles instructions into structured requirements, a Sculptor executes local scene operations, and an Inspector verifies each step while updating Working, Scene, and Skill Memory. To evaluate requirement-level controllability and preservation-aware editing, we introduce AuthorBench, offering 145 constrained construction cases and a 1,584-case preservation-aware editing pool paired with external structured checks. On full construction cases, MUSE improves All-Goal success from 37.9 to 80.7 and surface-constraint fulfillment from 35.0 to 92.6 over the strongest baseline. On a stratified 240-case editing test split, MUSE achieves 49.6 All-Goal success, 99.9 preservation rate, and only 0.6 unintended change rate. Beyond automated metrics, human evaluations on compared local-editing baselines support stronger alignment with user intent, and downstream navigation-proxy tests indicate stronger spatial stability. Combined with ablations validating our memory designs, these results establish MUSE as an effective framework for controllable 3D scene authoring.

06.
arXiv (CS.CV) 2026-06-11

Moving Beyond Diffusion: Hierarchy-to-Hierarchy Autoregression for fMRI-to-Image Reconstruction

Reconstructing visual stimuli from fMRI signals is a central challenge bridging machine learning and neuroscience. Recent diffusion-based methods typically map fMRI activity to a single neural embedding, using it as static guidance throughout the entire generation process. However, this fixed guidance collapses hierarchical neural information and is misaligned with the stage-dependent demands of image reconstruction. In response, we propose MindHier, a coarse-to-fine fMRI-to-image reconstruction framework built on scale-wise autoregressive modeling. MindHier introduces three components: a Hierarchical fMRI Encoder to extract multi-level neural embeddings, a Hierarchy-to-Hierarchy Alignment scheme to enforce layer-wise correspondence with CLIP features, and a Scale-Aware Coarse-to-Fine Neural Guidance strategy to inject these embeddings into autoregression at matching scales. These designs make MindHier an efficient and cognitively aligned alternative to diffusion-based methods by enabling a hierarchical reconstruction process that synthesizes global semantics before refining local details, akin to human visual perception. Extensive experiments on the NSD dataset show that MindHier achieves superior semantic fidelity, 4.67$\times$ faster inference, and more deterministic results than the diffusion-based baselines.

07.
bioRxiv (Bioinfo) 2026-06-18

Elucidating the Design Space of Generative Models for Single-Cell Perturbation Prediction

Next-token prediction has produced predictable scaling in language, but the recipe presumes a sequence of tokens with a meaningful order. Single-cell RNA-seq counts have no natural gene ordering, so applying the recipe directly to raw expression fails under an ill-suited left-to-right bias. We instead ask whether a learned latent can supply the structure the recipe needs. We introduce texttt{ExpressionVAE} (eVAE), a discrete-latent perturbation model that compresses each cell into a short sequence of discrete codes through a finite-scalar-quantization (FSQ) bottleneck and trains a perturbation-conditioned discrete prior over those codes. On Replogle and Parse~1M, eVAE sets a new state of the art on every distributional metric and leads on most cell-eval perturbation metrics, with Fr'echet distance and $mathrm{MMD}^2$ roughly $3$ to $20times$ lower than the strongest continuous-latent baseline. Swapping the prior between autoregressive and masked discrete diffusion leaves performance near-identical, isolating the gain to the discrete latent itself rather than the prior family. A decoder-head ablation then exposes a single design axis, the richness of the predictive distribution at inference, that splits the standard metrics into two groups, variance-sensitive and mean-sensitive, which move in opposite directions along the axis. Finally, on a held-out CRISPRi reversion benchmark of $1{,}732$ perturbations under inflammatory cytokine stress, the frozen eVAE encoder outperforms UMAP and differential expression and matches scGPT on perturbation ranking at a fraction of the data.

08.
arXiv (quant-ph) 2026-06-16

The Distribution Postulate in Algorithmic Bohmian Mechanics

arXiv:2606.16165v1 Announce Type: new Abstract: In order to make the right empirical predictions Bohmian mechanics requires a special statistical boundary condition – the distribution postulate – but it is unclear how best to understand this condition. We show how one might use the theory of algorithmic randomness to formulate the distribution postulate as an objective constraining law. The framework requires us to say something about admissible quantum-mechanical states and measurements. In return, algorithmic Bohmian mechanics (aBM) guarantees the standard Born statistics for a collection of canonical quantum experiments in the limit, not just with high probability. The algorithmic distribution postulate provides a sharp typicality condition, clarifies the status of quantum probabilities in the deterministic theory, and provides a concrete example of how notions provided by the theory of algorithmic randomness can aid in specifying the content of a physical law.

09.
arXiv (math.PR) 2026-06-17

Optimal Impulse Control for Cyber Risk Management

arXiv:2410.17706v2 Announce Type: replace-cross Abstract: We explore an optimal impulse control problem wherein an electronic device owner strategically calibrates protection levels against cyber attacks. Utilizing epidemiological compartment models, we qualitatively characterize the dynamics of cyber attacks within the network. We determine the optimal protective measures against effective hacking by formulating and solving a stochastic control problem with optimal switching. We demonstrate that the value function for the cluster owner constitutes a viscosity solution to a system of coupled variational inequalities associated with a fully coupled reflected backward stochastic differential equation (BSDE). Furthermore, we devise a comprehensive algorithm alongside a verification procedure to ascertain the optimal timing for network protection across various cyber attack scenarios. Our findings are illustrated through numerical approximations employing deep Galerkin methods for partial differential equations (PDEs). We visualize the optimal protection strategies in the context of two distinct attack scenarios: (1) a constant cyber attack, (2) an exogenous cyber attack strategy modeled with a Poisson process.

10.
arXiv (CS.LG) 2026-06-12

Normative Robustness as a Frontier for Non-Verifiable Reasoning in LLMs

arXiv:2606.12731v1 Announce Type: new Abstract: As LLMs increasingly serve in advisory and deliberative roles, users rely on them for non-verifiable reasoning in domains lacking objective ground truths. However, traditional evaluations of LLM reasoning focus almost exclusively on fact-based domains, such as mathematics and science, leaving uncertainty over whether and to what degree models can handle ambiguous, subjective, or value-laden problems over time. To address this concern, we propose moral reasoning as a paradigmatic subdomain of non-verifiable reasoning. We define moral robustness as a model's capacity to exhibit sound moral reasoning across time and contexts, and we introduce a scalable, adversarial, multi-turn evaluation framework to empirically measure this capability. We simulate 48,000 user-agent moral deliberations across four frontier LLMs, varying premise relevance, premise order, conversation duration, and the user's stated moral view. We find that models successfully ignore morally-irrelevant distractors, but shift their reasoning by up to 6.5%, on average, towards the user's stated preferred moral view, and varying their reasoning depending on factors such as order (altering moral judgments by order in 13-22% of the cases) and duration (altering moral judgments between single-turn and multi-turn in 10-24% of the cases). Our analysis indicates that models tailor not just their final verdicts but their underlying justifications to align with a user's moral viewpoint - a failure mode we characterize as moral deliberative sycophancy.

11.
arXiv (CS.LG) 2026-06-16

ROVE: Unlocking Human Interventions for Humanoid Manipulation via Reinforcement Learning

arXiv:2606.17011v1 Announce Type: cross Abstract: Human interventions provide crucial corrective signals for post-training Vision-Language-Action (VLA) models. However, enabling seamless humanoid interventions is a formidable systems challenge due to complex whole-body kinematics and dexterous-hand control. Consequently, the collected intervention trajectories are often suboptimal, and methods that rely on human interventions as expert supervision can absorb hesitant, inefficient, or even erroneous behaviors. To address both the system and algorithmic challenges, we propose ROVE, a reinforcement learning framework for humanoid VLA post-training with imperfect human interventions. First, ROVE introduces a human-in-the-loop pipeline capable of collecting deployment and intervention data for humanoid manipulation. Second, it utilizes Optimistic Value Estimation (OVE) to prioritize high-value behaviors from mixed-quality trajectories. To further robustify value estimation, we incorporate cross-embodiment human experience videos to provide rich supervision for long-tailed failure and recovery modes. The resulting critic yields informative advantage signals, steering the VLA actor to focus on high-value behaviors rather than indiscriminately imitating all actions. On challenging real-world contact-rich and fine-grained humanoid manipulation tasks, ROVE outperforms experience-learning baselines and consistently improves across multiple rollout-intervention iterations.

12.
arXiv (CS.LG) 2026-06-18

MetaboNet-Bench: A Multi-modal Benchmark for Glucose Forecasting in Type 1 Diabetes

arXiv:2606.18640v1 Announce Type: new Abstract: Glucose forecasting algorithms are an important aspect of glycemic control management in type 1 diabetes. So far, the research community has developed numerous algorithms and models for forecasting. However, it is well-recognized that the lack of standardized model performance evaluation benchmarks makes fair comparison difficult and hinders further innovation, and thus benchmark standardization is in urgent need. Furthermore, many published glucose forecasting algorithms are limited to CGM data alone, ignoring other multimodal signals such as insulin dosing and carbohydrate intake. Here, we introduce MetaboNet-Bench, a benchmark for multimodal glucose forecasting for patients with type 1 diabetes that provides an extensible open-source evaluation framework for comparison of glucose forecasting algorithms that leverage glucose, insulin, and carbohydrate data. We then demonstrate its utility by benchmarking several recently published glucose forecasting models and a custom multimodal time-series model, representing different model architectures. The results show that the benefit of adding data modalities is conditioned on the complexity of the model and that incorporating more clinical metrics helps identify meaningful gaps to fill for future research.

13.
Nature (Science) 2026-06-10

A first-in-class pulsatile FXR agonist for bile-acid-related liver diseases

作者:

Nuclear receptors are central regulators of metabolism1, yet therapeutic strategies that enforce continuous receptor activation frequently lead to reduced efficacy and unacceptable toxicity. Here we report a first-principles drug design strategy that aligns pharmacokinetics with physiological signalling cycles. We developed linafexor, a potent non-bile-acid agonist of the farnesoid X receptor (FXR)2; it is engineered for rapid systemic clearance, which enables pulsatile receptor activation that mirrors endogenous bile acid dynamics3–5. Linafexor has robust efficacy across multiple preclinical models of metabolic dysfunction-associated steatohepatitis6, liver fibrosis7, primary biliary cholangitis and primary sclerosing cholangitis8,9. Transcriptomic analyses reveal that, unlike long-acting FXR agonists10,11, linafexor preserves cyclic FXR signalling, avoids receptor downregulation and prevents broad transcriptional dysregulation. Direct manipulation of delivery patterns demonstrates that sustained FXR activation—independent of compound identity—induces severe toxicity, establishing activation duration as a determinant of therapeutic index. In phase 1 clinical studies (ClinicalTrials.gov; NCT05082779), linafexor administered once daily produces transient FXR pathway engagement, marked by (1) induction of FGF1912–14, a key endocrine mediator of bile acid feedback regulation; and (2) suppression of C415, an intermediate reflecting hepatic bile acid synthesis, with no treatment-related adverse events. Together, these findings identify pulsatile FXR activation as a mechanistically grounded and clinically translatable strategy, and establish linafexor as a first-in-class therapeutic for bile acid–related liver diseases. Linafexor is a rapidly cleared FXR agonist designed to mimic natural bile acid signalling, achieving transient receptor activation with strong efficacy and reduced toxicity in preclinical and early clinical studies.

15.
arXiv (CS.AI) 2026-06-16

Bayesian Inference and Decision Audits for Public Archives of Frontier AI Evaluations

作者:

arXiv:2606.17005v1 Announce Type: new Abstract: Public AI evaluations are often read as terminal leaderboards, yet the underlying evidence is a selective time series shaped by reporting rules, benchmark revisions, and missingness. Repeated public archives for LiveBench and Open LLM Leaderboard v2 serve as the primary longitudinal record; LMArena provides a preference stress test; and GAIA and tau-bench contribute limited agentic pilots. Together, these archives instantiate a Bayesian inference problem: under a fixed reporting convention, one constructed terminal-only example over $1{,}000$ systems is compatible with two pre-terminal histories, yielding times of $23.03$ or $75.13$ to reach within $0.05$ of the ceiling under the same terminal-tail model. In synthetic posterior comparisons, action-facing diagnostics differ across observation regimes. The candidate selection-aware frontier model fails synthetic recovery, objective-archive prediction, preference transfer, and uncertainty calibration; correspondingly, fixed audit gates reject its stronger claims. An archive-and-adjudication protocol reconstructs public evaluation histories, isolates a verified timing boundary, and falsifies unsupported frontier claims.

16.
arXiv (CS.AI) 2026-06-15

When Sample Selection Bias Precipitates Model Collapse

arXiv:2606.13732v1 Announce Type: new Abstract: The proliferation of recursive training on synthetic data can alleviate data scarcity but risks model collapse, where repeated training erodes distributional tails and homogenizes outputs. Data selection is widely viewed as a remedy, yet its reliability depends critically on the reference distribution used by the verifier. We show that in low-resource verification regimes, where each verifier observes only a small, fragmented, and biased slice of the target manifold, selection itself becomes biased. This situation naturally arises in low-resource data silos such as healthcare consortia or proprietary financial institutions, where raw data cannot be pooled and local references are inherently incomplete. As a result, selection preferentially retains samples aligned with the local manifold while pruning globally relevant tail modes, turning from a safeguard against collapse into a mechanism that precipitates it. We theoretically prove that such siloed selection accelerates collapse and induces power-law diversity decay. As an initial mitigation, we construct Wasserstein proxy references from multiple silos without sharing raw data. Empirical results confirm that local-reference selection fails on skewed distributions, whereas collaborative proxy references mitigate diversity degradation, suggesting that recursive synthetic-data pipelines require particular caution when real-data coverage is fragmented or scarce.

17.
arXiv (CS.CL) 2026-06-11

Fine-tuning Multi-modal LLMs with ART: Art-based Reinforcement Training

There are two main Parameter-Efficient Fine-Tuning (PEFT) techniques for Large Language Models (LLMs). While Low-Rank Adaptation (LoRA) introduces additional weights between the LLM layers, Soft Prompting introduces additional fine-tuning-specific raw tokens to an LLM input. However, both require modification to the computational graphs of precompiled, preoptimized LLMs. As a result, neither is fully supported in high-throughput engines like vLLM. We propose fine-tuning with ART (Art-based Reinforcement Training). The method injects information into a frozen Multimodal Large Language Model (MLLM) by optimizing only its raw visual input, thus enabling the soft-token approach on pre-compiled computational graphs. It relies on backpropagation of gradients back into a plain pixel array and thus supports any fine-tuning objective. Moreover, the optimized visual input can be stylized as task-relevant computational artworks. The approach's effectiveness is confirmed for different sizes of a popular open Qwen architecture and for several textual benchmarks. Specifically, ART reaches accuracy competitive with LoRA across mathematics and structured-tool-use benchmarks.

18.
arXiv (math.PR) 2026-06-18

Law of the Iterated Logarithm for $p$-Walks on $\mathbb{Z}$

作者:

arXiv:2606.19131v1 Announce Type: new Abstract: The $p$-rotor walk on $\mathbb{Z}$ is a self-interacting walk that interpolates between the simple random walk and the deterministic rotor walk. While the weak convergence of this model to a perturbed Brownian motion is known, its almost sure asymptotic boundaries have not been characterized. In this paper, we establish the exact Law of the Iterated Logarithm (LIL) for the $p$-rotor walk. Utilizing the decomposition of the walk into a martingale perturbed by its running extrema, we obtain first a functional Law of the Iterated Logarithm for the linearly interpolated paths of the $p$-walk. We then obtain the classical LIL constants by solving a calculus of variations problem over the perturbed Strassen set.

19.
arXiv (math.PR) 2026-06-16

Universality in the target arrival statistics of non-conservative search processes

arXiv:2606.16025v1 Announce Type: cross Abstract: Stochastic search processes in which searchers are continuously introduced to and removed from a target search domain are fundamental to a wide class of physical and artificial systems. The theory of such non-conservative search processes is, however, much less developed than for search processes with a fixed number of particles. Here we exploit a natural mapping between non-conservative stochastic search and queueing theory to derive the full time-dependent distribution of target arrivals under minimal assumptions on the underlying search process. Remarkably, we find that the steady-state inter-arrival time distribution is exactly exponential, regardless of the details of the search process, showing a robust universality that emerges directly from the queueing framework. Thus, counterintuitively, the arrival statistics of a non-conservative search process are much simpler than sequential search-and-capture processes involving a fixed number of searchers. This has major implications for target resource accumulation, where the delivery of resources is counter-balanced by their downstream consumption.

21.
arXiv (CS.CL) 2026-06-19

Generative Engine Optimization at Scale: Measuring Brand Visibility Across AI Search Engines

People increasingly get answers straight from AI search engines like ChatGPT, Claude, Perplexity, and Gemini rather than scrolling search results. Brands that once focused on search engine optimization (SEO) must now optimize for how these engines represent, cite, and recommend them – a shift variously called Generative Engine Optimization (GEO), Answer Engine Optimization (AEO), and AI Search Visibility. We treat AEO and AI Visibility as part of GEO, and study how to measure brand visibility across AI engines: what they value when they cite a brand, which sources they rely on, and what content large language models surface. The hard case is everyone outside the already-authoritative top brands – SMEs, D2C brands, creators, and early-stage startups. We analyze 100K+ prompt responses across 100+ brands tracked on Ranqo between March and May 2026. First visibility runs form a clear three-tier brand-stature ladder: global household names (e.g., Stripe, Nike) appear in 73% of relevant AI answers on their first run; established mid-market and regional brands (e.g., Olipop, Klaviyo) in 44%; niche and small brands in just 11% – about 30 percentage points per step. When engines cite sources, about 78% go to corporate websites; among non-corporate sources YouTube leads, ahead of Reddit, editorial media, and Wikipedia. The highest-leverage page is the ranked "best-of" listicle, the most-cited content format at about 21% of all citations. Sentiment is the unstable signal: whether a brand is framed positively or negatively flips about 6.7 times more often than whether it is mentioned at all. These findings provide a first large-scale baseline for measuring GEO: AI brand visibility can be measured, differs by platform, and varies strongly by brand maturity. We close by proposing seven v1.1 protocols to test whether specific recommendations can causally improve AI visibility.

22.
Nature (Science) 2026-06-17

Towards autonomous medical artificial intelligence agents

作者:

Large language models (LLMs) show great potential for clinical decision-making, yet most applications remain narrow, task-specific chat tools rather than systems integrated into clinical workflows1,2. However, building physician copilots will require models that operate within the electronic health record (EHR), with governed access to patient data and the ability to initiate permitted EHR actions within defined safety constraints. Yet it remains unproven whether such a system can manage patient cases with physician-level performance. Here we show that MIRA (Medical Intelligence for Reasoning and Action), an autonomous artificial intelligence agent operating in a sandboxed EHR environment, can navigate a large clinical action space to obtain patient histories; order and interpret laboratory, imaging and microbiology tests; generate differential diagnoses; and formulate treatment plans such as prescribing medications, scheduling surgical procedures and planning admissions. In simulations on real patient cases spanning multiple diagnoses, MIRA outperformed physicians in diagnostic accuracy and made guideline-concordant, medication-safe and appropriate admission decisions. Compared with previous LLM applications that addressed isolated subtasks or provided free-text advice, these results suggest that an EHR-integrated artificial intelligence agent can turn clinical intent into structured, actionable EHR operations, possibly making it a more effective decision-support partner for physicians. Further work is needed to establish generalization, safety and governance through prospective, real-world studies. A large language model artificial intelligence agent operating in a sandboxed electronic health record system can autonomously take patient histories, order tests, interpret findings, diagnose conditions and propose treatments, outperforming experienced clinicians while adhering to safety standards and clinical guidelines.

23.
arXiv (CS.AI) 2026-06-11

Generalization Hacking: Models Can Game Reinforcement Learning by Preventing Behavioral Generalization

arXiv:2606.12016v1 Announce Type: cross Abstract: Model post-training, and in particular reinforcement learning (RL), is one of the primary mechanisms by which developers can shape models' values and behaviors. However, as models become increasingly evaluation and training aware, they may be motivated to resist training when the perceived objective conflicts with their current values, undermining developers' ability to detect misalignment and correct model behavior through further training. In this paper, we demonstrate generalization hacking, in which a model collects reward during RL while preventing the rewarded behavior from generalizing. We construct a model organism on Qwen3-235B-A22B, finetuning on synthetic documents describing training awareness and self-inoculation, a novel mechanism in which the model frames compliance as context-specific in its chain of thought, without demonstrating or instructing either behavior. The model organism achieves train-time harmfulness comparable to controls while maintaining a persistent ${\sim}15$ percentage point compliance gap across 700 steps of RL. Additionally, a control organism trained only on training awareness documents independently discovers inoculation-like reasoning under RL pressure, developing its own compliance gap despite never being exposed to the concept. Because the generalization-hacking organism receives high reward throughout, standard training metrics provide no signal that generalization has failed. Our results constitute the first demonstration that a model can actively resist RL behavioral modification while maintaining high reward, suggesting that as models become more capable and training-aware, they may be able to undermine the training process itself.

24.
arXiv (CS.AI) 2026-06-16

HoloRec: Holistic Encoding and Interleaved Reasoning for Generative Recommendation

arXiv:2606.15331v1 Announce Type: cross Abstract: Generative recommendation models that formulate the task as sequence generation overcome the objective fragmentation problem of traditional cascade architectures, yet existing approaches still suffer from flat semantic representations lacking hierarchical structure for multi-step reasoning and an externally constructed chain-of-thought (CoT) that requires expensive annotations and remains disconnected from the generation objective. We propose HoloRec, an endogenous chain-of-thought recommendation mechanism that unifies representation, reasoning, and generation by constructing a hierarchical semantic encoding matrix via multi-granularity nested residual quantization optimized by a holistic reconstruction loss. HoloRec supports two inference modes: a non-thinking mode that uses lightweight multi-granularity supervised alignment for fast prediction, and a thinking mode that employs an interleaved reasoning scheme to generate CoT steps on the fly, directly embedding reasoning into the generation process without external data. Experiments on multiple public recommendation datasets demonstrate that HoloRec consistently outperforms baselines, with especially significant gains in sparse scenarios, and the thinking mode achieves better accuracy than the non-thinking mode with only modest inference overhead.

25.
arXiv (CS.AI) 2026-06-11

Steering Where to Listen: Instruction-Based Activation Steering Redirects Temporal Attention in Large Audio-Language Models

arXiv:2606.11400v1 Announce Type: cross Abstract: Large Audio-Language Models (LALMs) excel at audio understanding but expose little about where in an audio signal they attend. We introduce instruction-based vector steering, which constructs a steering vector by contrasting activations from differently instructed prompts while keeping the audio fixed. Through a systematic probe of LALM attention, we find that - unlike standard prompting or audio-based steering - this intervention significantly redistributes the temporal attention allocated to audio tokens, concentrating it on acoustically relevant regions. We then show that this attention shift is behaviorally meaningful: in a controlled three-event setting, reading out the temporal position of maximal steering-induced attention change recovers the location of a queried sound event without any training, attaining 60.87% and 68.72% overlap with ground-truth intervals on Qwen2-Audio and Audio Flamingo 3, far above direct prompting (31.84%, 46.75%) and random baselines (27.74%). Our results characterize a mechanistic property of instruction-based steering in LALMs and provide a training-free probe for the latent temporal structure these models encode.