arXiv:2607.06192v1 Announce Type: new Abstract: Virtual admittance (VA) is widely used in cascaded voltage-control and current-control (VC-CC) grid-forming inverters (GFMIs) because it shapes the converter terminal behavior while preserving the current-regulation path required for current shaping and limiting. However, the achievable VC-loop bandwidth remains strongly coupled to the CC-loop bandwidth and to the VA parameters. Voltage-feedback decoupling (VFD) is commonly used to relax this coupling, but in VA-based control its benefit is not unconditional. This paper shows that unity-gain VFD, which represents the full-decoupling condition, removes the low-frequency restoring term associated with the filter capacitor and drives the voltage loop toward a delay-sensitive double-integrator structure. This internal-stability limitation is referred to here as the VFD trap. To address this trap without attenuating VFD, a proportional active-damping (AD) path is proposed, implemented as negative capacitor-voltage feedback in the current-reference path. The proposed path restores the missing low-frequency support while retaining unity VFD and introduces an additional AD-based degree of freedom for VC-loop tuning. A minimum support condition, a delay-aware phase-margin expression, and compact forward/inverse design equations are derived for operating-point selection. Standalone and grid-connected experiments on a 3-kVA prototype verify the analysis, showing that the proposed path recovers stable unity-VFD operation, reduces the voltage-step settling time from approximately 9~ms to 3~ms, and maintains stable power injection.
Science Journals
arXiv:2607.05868v1 Announce Type: new Abstract: Designing cost functions of adaptive steganography traditionally requires extensive manual tuning, while deep learning methods lack interpretability. Although large language models (LLMs) offer an automated alternative via evolutionary generation, they often violate domain specific mathematical constraints due to a lack of explicit domain knowledge. To address this problem, we propose a novel evolutionary system focused on exploiting Retrieval-Augmented Generation (RAG) enhanced LLMs for the automatic code-level generation of spatial steganography cost functions. This system incorporates a core Self Evolving RAG (SE-RAG) module, wherein a Code Semantic Signature (CSS) translates procedural code into aligned queries, retrieving explicit guidance from static literature and dynamic experience knowledge bases to steer the LLM generation process. A dedicated feedback mechanism then continuously refines the dynamic knowledge base with successful optimization strategies. Extensive experiments on the BOSSBase and BOWS2 datasets demonstrate that the proposed framework consistently achieves higher steganographic security than existing automatically designed methods, and increases the average code execution rate by 46.3% while reducing the search cost by 26.1%, thereby highlighting the effectiveness, efficiency, and potential of combining LLMs with domain-specific knowledge in the field of automatic steganographic algorithm generation.
arXiv:2211.10942v2 Announce Type: replace-cross Abstract: Difference-of-Convex (DC) programming, which seeks to minimize a function expressed as the difference of two convex functions, arises in a wide range of applications in machine learning, signal processing, and operations research. A classical and widely used algorithm for solving DC programs is the Difference-of-Convex Algorithm (DCA). In this paper, we revisit DCA from a distinctly DC-specific perspective. We first separate well-definedness from asymptotic convergence and introduce an additional assumption ensuring the solvability of the DCA subproblems, which clarifies why the choice of DC decomposition matters. We then develop a Lyapunov-descent-regularity framework in which the descent estimate is read directly from the convex subproblems and the regularity estimate is verified from DCA optimality conditions. This yields global convergence of the iterates $\{x^k\}$ for both standard and convex-constrained DC programs under either the classical Lojasiewicz subgradient inequality or the broader Kurdyka-Lojasiewicz (KL) property. We further explain how stronger regularity regimes, such as the Polyak-Lojasiewicz (PL) condition, fit into the same framework and sharpen the resulting convergence rates. Consequently, we obtain finite-time, linear, and sublinear rates for objective values and iterates in a way that cleanly separates well-definedness, DCA-specific structure, and KL/PL regularity, and that is readily transferable to DCA-type variants.
arXiv:2512.00517v3 Announce Type: replace-cross Abstract: Sequential optimization of black-box functions from noisy evaluations has been widely studied, with Gaussian Process bandit algorithms such as GP-UCB guaranteeing no-regret in stationary settings. However, for time-varying objectives, no-regret is unattainable under pure bandit feedback unless strong and often unrealistic assumptions are imposed. We propose a novel method for optimizing time-varying rewards in the frequentist setting, where the objective has bounded RKHS norm almost surely. Time variations are captured through uncertainty injection, enabling heteroscedastic Gaussian process regression that adapts past observations to the current time step. As no-regret is unattainable in general in the strict bandit setting, we relax the latter allowing additional queries on previously observed points. Building on sparse inference and the effect of uncertainty injection on regret, we propose W-SparQ-GP-UCB, an online algorithm that achieves no-regret with a vanishing number of additional queries per iteration. To assess the theoretical limits of this approach, we establish a lower bound on the number of additional queries required for no-regret, proving the efficiency of our method. Finally, we provide a comprehensive analysis linking the temporal regime of the function to achievable regret rates, together with upper and lower bounds on the number of additional queries needed in each regime.
arXiv:2607.06120v1 Announce Type: new Abstract: Text-to-image diffusion models have achieved high visual fidelity and broad adoption, but remain vulnerable to safety violations when adversaries exploit them to synthesize illicit content. Existing alignment paradigms, from input sanitization to structural feature pruning, are largely organized around unsafe concepts explicitly exposed during filtering, editing, or localization. This leaves a blind spot for visual synonym attacks (VSA), a jailbreak where benign-looking prompts elicit prohibited imagery through implicit visual associations. As a result, current defenses face a safety-utility dilemma: they may either under-mitigate VSA threats or over-suppress visually similar benign concepts. The core challenge is that VSA hides the unsafe target at the textual surface while revealing it through generation-time visual-semantic convergence. In this work, we therefore shift from static suppression of pre-specified unsafe concepts to dynamic tracing of how unsafe semantics emerge during generation. Our mechanistic analysis shows that VSA and explicit unsafe prompts converge through sparse semantic-injecting attention heads, which serve as inference-time bottlenecks for prohibited visual semantics. Based on this insight, we propose AEGIS (Adaptive Evasion Guard via Identification and Steering), an inference-time defense that applies similarity-aware repulsion only at the identified vulnerable heads. Evaluated against 16 baselines, AEGIS improves both safety and utility. On SD 1.4, it reduces ASR to $\mathbf{0.00}/\mathbf{0.03}$ for in-domain violence/nudity VSA and achieves ASRs $\le \mathbf{0.09}$ on out-of-domain explicit and adversarial attacks. It preserves benign fidelity, avoids suppressing hard-negative concepts, and transfers to SD 2.1 and FLUX.1 after re-identifying the critical heads for each backbone.
arXiv:2607.05559v1 Announce Type: new Abstract: Foundation machine learning force fields (MLFFs) such as MACE-MP-0 and UMA cover broad chemical space at near density functional theory (DFT) accuracy. However, they assume equilibrium ground-state physics and do not natively handle externally induced changes to the electronic state, such as charging, applied fields, or electronic excitation, which limits their use for driven processes such as photoexcitation and charge injection. We propose EquiFiLM, a lightweight extension that adds continuous external conditioning to any equivariant foundation MLFF via a per-layer Feature-wise Linear Modulation (FiLM) block, learning externally driven changes to the potential energy surface from minimal training data. The block modulates only scalar channels and preserves E(3)-equivariance exactly. We demonstrate the recipe on charged liquid water with the foundation model MACE-MatPES as the backbone, yielding E-MACE. On the four training charges, E-MACE delivers a $3.1\times$ reduction in force RMSE ($21.3$ to $6.96$ meV/$\mathring{A}$) and a $61\times$ reduction in per-atom energy RMSE ($6.1$ to $0.1$ meV/atom) over a baseline without EquiFiLM trained on the same data, at indistinguishable inference cost. Across seven held-out interpolation and extrapolation charges, force RMSE stays within $18-61$ meV/$\mathring{A}$ and energy RMSE within $0.7-5.4$ meV/atom. The model runs stable molecular dynamics across the full range tested and predicts the charge-dependent first-shell response of the reduced pair distribution function probed by ultrafast electron diffraction. Adding this conditioning axis to the foundation requires only a few thousand DFT-labeled frames, against the $\approx 10^8$ structures of a charge-aware foundation trained from scratch. The recipe is backbone- and conditioning-agnostic: it applies without architectural change to any equivariant MLFF with scalar interaction-layer channels.
arXiv:2607.06392v1 Announce Type: new Abstract: Self-supervised learning (SSL) models, such as Wav2Vec2, HuBERT, and WavLM, have become foundational across a wide range of speech and audio tasks. Despite their success, understanding their internal layer-wise dynamics remains an ongoing challenge. To address this, we propose a two-part model-centric framework called InsideSSL. First, we establish a task-agnostic analysis from three intrinsic per-layer perspectives: compression (entropy), geometry (curvature), and robustness to perturbations. We show that varying training objectives induce distinct regimes of acoustic compression and manifold unfolding. Second, we introduce the cross-layer Generative Compatibility Matrix (GCM) to evaluate functional transferability, exposing stable phonetic cores, identity volatility, and deep-layer semantic pruning. In addition to these evaluations, linear probing connects the model-centric perspective to downstream tasks, demonstrating how layer topology dictates phoneme, pitch, and speaker encoding.
arXiv:2607.04415v2 Announce Type: replace Abstract: Graph-based approximate nearest neighbor search (ANNS) is increasingly used in vector databases and retrieval-augmented generation services, but large vector indexes often exceed the memory capacity of a single GPU server. NVMe over Fabrics (NVMe-oF) provides an attractive storage-disaggregation substrate, yet existing remote storage paths are still largely CPU-centered: the CPU forms I/O requests, drives transport progress, and determines when GPU computation can resume. This organization is poorly matched to graph ANNS, where the next data access is discovered inside GPU graph traversal. This paper presents GORIO, a system study that extends GPU-centered local I/O to remote storage and specializes the resulting substrate for graph ANNS over NVMe-oF. GORIO keeps query evolution, page-miss generation, pending-query state, and resume decisions on the GPU, while the CPU acts only as an NVMe-oF transport and completion proxy. The design has two layers: a GPU-direct remote I/O path that turns local page-cache misses into split-phase remote operations, and ANNS-specific scheduling mechanisms that overlap graph traversal with remote page service. On a SIFT1M DiskANN-style graph workload over an RDMA NVMe-oF path, GORIO is 1.31X faster than the state-of-the-art remote-I/O reference path and 4.89X faster than the direct remote page-cache path. These results demonstrate a concrete GPU-centered remote I/O substrate for graph ANNS.
arXiv:2603.12995v4 Announce Type: replace-cross Abstract: The subtour relaxation of the traveling salesman problem (TSP) plays a central role in approximation algorithms and polyhedral studies of the TSP. A long-standing conjecture asserts that the integrality gap of the subtour relaxation for the metric TSP is exactly 4/3. In this paper, we extend the exact verification of this conjecture for small numbers of vertices. Using the framework introduced by Benoit and Boyd in 2008, we confirm their results up to n=10. We further show that for n=11 and n=12, the published lists of extreme points of the subtour polytope are incomplete: one extreme point is missing for n=11 and twenty-two extreme points are missing for n=12. We extend the enumeration of the extreme points of the subtour polytope to instances with up to 15 vertices in the general case. Restricted to half-integral vertices, we extend the enumeration of extreme points up to n=17. Our results provide additional support for the 4/3-Conjecture. Our lists of extreme points are available on the public bonndata repository (https://doi.org/10.60507/FK2/JK95PC).
arXiv:2607.05450v1 Announce Type: new Abstract: This paper explores the "Granularity Paradox" in time-series forecasting, wherein finer temporal disaggregation (e.g., Monthly to Weekly/Daily) improves in-sample diagnostics and dataset size (N), but degrades out-of-sample accuracy due to recursive error compounding over longer horizons (H). Conversely, coarse aggregation (Annual) eliminates recursive error propagation but reduces data available to estimators. We formalize this trade-off and benchmark 10 models - spanning na\"ive, statistical, machine learning, and deep learning architectures - across six granularities using a 13-year public procurement dataset. The empirical results reveal a non-monotonic threshold structure: recursive autoregressive and seasonal models degrade substantially under high-frequency forecasting (e.g., Holt-Winters reaches a Test R-squared of -151 and TPFE of 425.85% at the Daily grain), while the LSTM traces a U-shaped error curve, worsening from Monthly (19.66%) through Bi-Weekly (35.94%) before overcoming the error propagation penalty at Daily (TPFE of 4.35%, R-squared of 0.66). Linear Regression remains stable across all granularities (16.3-17.0% TPFE), confirming that the paradox is driven by recursive feedback topology, not model complexity. The results demonstrate that standard pointwise metrics (RMSE, MAE) systematically mask cumulative error propagation, and that evaluating forecasts without goal-dependent cumulative metrics produces misleading assessments of model adequacy. We introduce a consensus-dissensus diagnostic comparing the directional behaviour of pointwise metrics against cumulative TPFE across granularities, enabling the identification of models whose standard diagnostics mask systematic error propagation.
arXiv:2607.06453v1 Announce Type: new Abstract: Low-rank tensor methods are an important tool in the numerical treatment of equations with a high-dimensional state space. Nearest neighbor interaction systems like the Ising model or more general Markov jump processes, as well as 1D finite-state quantum systems are examples of such problems. While low-rank tensor train/matrix product state models have been shown to be highly efficient for the simulation of such systems, providing theoretical justification for this remains a challenging task. One approach for obtaining estimates on required ranks for certain accuracies is to investigate the rank increase in Krylov subspace methods for solving the problem at hand. In the context of area laws for ground states of 1D spin systems, nontrivial results on rank-increasing properties of nearest neighbor operator polynomials have been obtained in work of Arad et al. [arXiv:1301.1162] by studying the partial commutativity of local operators. In the present work, this technique is applied to polynomial methods for definite linear equations and dissipative linear ODEs with nearest neighbor structure. This allows to derive corresponding low-rank approximability statements for solutions of such problems which are independent of the system size. Numerical simulations of high-dimensional nearest neighbor systems illustrate the theoretical findings.
arXiv:2607.06458v1 Announce Type: new Abstract: The analytical modeling of a plasma wakefield driven by a relativistic proton beam is an element in optimizing advanced plasma-based acceleration schemes. In this work, we present a 1D nonlinear fluid framework under the quasi-static approximation to describe the wake potential excited by a positively charged proton driver. We examine our model using a two-bunch pump-probe configuration, demonstrating close agreement between the analytical invariants and adaptive numerical integrations. The distinct geometric curvature changes observed at the micro-bunch boundaries are shown to be physical consequences of step-discontinuities in the second derivative of the wake potential across the beam interfaces. Furthermore, by scaling this numerical framework to a train of $N=100$ micro-bunches undergoing seeded self-modulation (SSM), we model the physical parameters of the CERN AWAKE facility ($n_0 = 7.0 \times 10^{14}\text{ cm}^{-3}$). Our model replicates the characteristic linear growth envelope and matches the calibrated field envelope boundaries of approximately $\pm 0.75\text{ GV/m}$ inferred from the experiment. This piece-wise framework provides a computationally efficient foundation for investigating customized, asymmetric micro-bunch profiles designed to optimize the transformer ratio beyond the fundamental symmetric limit of 2.
arXiv:2607.04837v2 Announce Type: replace Abstract: Large-scale humanoid motion-tracking controllers are commonly improved by reallocating training effort: difficult motions are sampled more often, isolated into smaller subsets, or assigned to specialized experts. We show that this view is incomplete. In strong whole-body-control baselines, a residual set of feasible training clips remains unsolved even under targeted training, especially for high-dynamic transitions and balance-critical motions. These failures arise not only from insufficient exposure, but from a mismatch between the motion demands and the effective capability induced by the default training recipe. We propose Athena-WBC, a compact teacher-student pipeline with capability-aligned policy experts for long-tail humanoid whole-body control. Dynamic experts use a tracking-focused, constraint-aware objective that removes conservative effort and temporal-control penalties while preserving physical feasibility constraints; balance experts use a gravity curriculum to improve early-training survivability. The resulting privileged teachers are motion-routed for DAgger distillation and then compressed into a single controller with deployable observations followed by RL fine-tuning. Experiments on a full-size humanoid show improved recovery of training-set long-tail motions and better held-out tracking than a strong SONIC-recipe baseline, using only a small number of experts.
arXiv:2508.20735v2 Announce Type: replace Abstract: We study parallel algorithms for the minimisation and equivalence checking of Deterministic Finite Automata (DFAs). Regarding DFA minimisation, we implement four different massively parallel algorithms on Graphics Processing Units~(GPUs). Our results confirm the expectations that the algorithm with the theoretically best time complexity is not practically suitable to run on GPUs due to the large amount of resources needed. We empirically verify that parallel partition refinement algorithms from the literature perform better in practice, even though their time complexity is worse. Furthermore, we introduce a novel algorithm based on partition refinement with an extra parallel partial transitive closure step and show that on specific benchmarks it has better run-time complexity and performs better in practice. In addition, we address checking the language equivalence and inclusion of two DFAs. We consider the Hopcroft-Karp algorithm, and explain how a variant of it can be parallelised for GPUs. We note that these problems can be encoded for the GPU-accelerated model checker \GPUexplore, allowing the use its lockless hash table and fine-grained parallel work distribution mechanism.
arXiv:2601.21738v2 Announce Type: replace Abstract: Evaluation of Image Quality Assessment (IQA) models has long been dominated by global correlation metrics, such as Pearson Linear Correlation Coefficient (PLCC) and Spearman Rank-Order Correlation Coefficient (SRCC). While widely adopted, these metrics reduce performance to a single scalar, failing to capture how ranking consistency varies across the local quality spectrum. For example, two IQA models may achieve identical SRCC values, yet one ranks high-quality images (related to high Mean Opinion Score, MOS) more reliably, while the other better discriminates image pairs with small quality/MOS differences (related to $|\Delta$MOS$|$). Such complementary behaviors are invisible under global metrics. Moreover, SRCC and PLCC are sensitive to test-sample quality distributions, yielding unstable comparisons across test sets. To address these limitations, we propose \textbf{Granularity-Modulated Correlation (GMC)}, which provides a structured, fine-grained analysis of IQA performance. GMC includes: (1) a \textbf{Granularity Modulator} that applies Gaussian-weighted correlations conditioned on absolute MOS values and pairwise MOS differences ($|\Delta$MOS$|$) to examine local performance variations, and (2) a \textbf{Distribution Regulator} that regularizes correlations to mitigate biases from non-uniform quality distributions. The resulting \textbf{correlation surface} maps correlation values as a joint function of MOS and $|\Delta$MOS$|$, providing a 3D representation of IQA performance. Experiments on standard benchmarks show that GMC reveals performance characteristics invisible to scalar metrics, offering a more informative and reliable paradigm for analyzing, comparing, and deploying IQA models. Codes are available at https://github.com/Dniaaa/GMC.
arXiv:2607.06522v1 Announce Type: new Abstract: Vision-language models (VLMs) struggle to generalize in interactive physical reasoning, particularly under unseen tasks and environments. Two key failure modes are prominent: hallucinated chain-of-thought (CoT) reasoning that contradicts physical reality, and misalignment between the model's reasoning and actions. We present VAORA (Visual Action Outcome Reasoning Alignment), a novel reward design that directly addresses both issues. VAORA introduces two complementary rewards: Visual Alignment Reward, which anchors VLM reasoning to the visual context independent of the agent action itself, and Visual-Action Alignment Reward, which grounds reasoning in the visual outcome induced by the model's action. Together, these rewards suppress hallucinated CoT and reduce the gap between reasoning and behavior. To improve training stability, we further employ smooth, dense rewards by estimating success probabilities using a pre-trained in-domain expert agent. Experiments on PHYRE and Virtual Tool support our performances across novel-task and unseen-environment settings, confirming that grounded and generalizable physical intelligence can be induced through VAORA.
arXiv:2607.06171v1 Announce Type: cross Abstract: Let $H$ be an inclusion-free hypergraph on $n$ vertices. A weight assignment $w:[n]\to[d]$ is isolating if there is a unique edge $e$ whose weight $w(e) = \sum_{i \in e} w(i)$ is minimum. We show that the number of isolating weight assignments is at least $$ n\sum_{j=0}^{d-1} j^{n-1}, $$ a bound which is attained with equality by the hypergraph consisting of the $n$ singleton edges. This proves the conjecture stated in Faber & Harris (2018). We also prove the bound for a more general class of edge-weight objectives, including arbitrary edge offsets.
arXiv:2503.09460v2 Announce Type: replace Abstract: The draft candidate European Cybersecurity Certification Scheme for Cloud Services (EUCS) defines security controls that must be associated with measurable metrics to assess compliance. This association process is currently manual, time-consuming, and prone to inconsistencies. In this paper, we propose an automated approach based on Sentence Transformers to associate cloud security controls with quantifiable metrics by leveraging semantic similarity between their textual descriptions. We evaluate our method on a dataset of 70 controls derived from the EUCS framework. The proposed approach outperforms a FastText-based baseline, achieving a conditional Normalized Discounted Cumulative Gain at rank 10 score of 0.640 (+0.146) and improving the standard nDCG@10 score from 0.275 to 0.504. These results demonstrate that contextual embedding models significantly enhance both the likelihood of retrieving relevant metrics and their ranking quality. Our findings highlight the potential of transformer-based methods to support automated, scalable, and more reliable compliance processes in cloud cybersecurity certification.
arXiv:2607.03820v2 Announce Type: replace-cross Abstract: Electromagnetically induced transparency suppresses optical absorption through destructive interference, playing a central role in light-matter interaction and quantum information science. We report matter-wave induced transparency, where atomic collisional interactions induce transmission through a lossy molecular potential for the incident atomic scattering waves. Using cesium Bose-Einstein condensates and modulation-induced Feshbach resonances, we realize a three-level atom-molecule coupled system with unprecedented flexibility. Under the dark state condition, a narrow and tunable transparency window appears within a broad dissipative collisional resonance. The transparency window linewidth is controlled by modulation-induced coupling. And scattering pathways are selectable via multifrequency Floquet modulation. These results establish an interference-based route for exploring programmable nonequilibrium and non-Hermitian physics, steering quantum chemistry and precision measurements.
arXiv:2607.06092v1 Announce Type: new Abstract: G.\ D.\ Baker formulated a forcing method to interpret integer optimisation problem into $2$-adic linear regression, and proved the NP-hardness of $2$-adic linear regression. We generalise the forcing method to a wider class of $p$-adic optimisation for the case where $p$ is not necessarily $2$, and prove the NP-hardness of $p$-adic linear regression, the NP-hardness of $2$-adic dynamic neural network by S.\ Albeverio, A.\ Khrennikov, and B.\ Tirrozi, and the NP-hardness of a partial generalisation of the $p$-adic optimisation problem associated to van der Put neural network by G.\ L.\ R.\ N'guessan.
arXiv:2607.05418v1 Announce Type: cross Abstract: In this paper, we study the group randomness of 0-1 real sequences derived from a binary linear code by investigating the spectral behaviour of a suitable normalization of the Gram matrix of a $p \times n$ random matrix whose rows are uniformly drawn from those 0-1 real sequences, where $y=p/n \in (0,1)$ is fixed. We show that as $n \to \infty$, its empirical spectral distribution converges to the Marchenko-Pastur law at a rate at least of the order $n^{-1/4}$ with high probability, and the fluctuation of its largest eigenvalue is asymptotically Gaussian with mean $p+1+y$ and variance $4y$, provided that the dual distance of the code is at least 5.
arXiv:2607.05459v1 Announce Type: cross Abstract: Tractatus Quanticum (arXiv:2512.06034 [quant-ph]) is described by its authors as 'a re-editing, which takes quantum mechanics into account, of Wittgenstein's famous Tractatus.' The original Tractatus appeared with an introduction by Bertrand Russell. For Tractatus Quanticum, that role fell to us. This is the result.
arXiv:2507.22758v2 Announce Type: replace Abstract: Recent advancements in financial problem-solving have leveraged LLMs and agent-based systems, with a primary focus on trading and financial modeling. However, credit assessment remains an underexplored challenge, traditionally dependent on rule-based methods and statistical models. In this paper, we introduce MASCA, an LLM-driven multi-agent system designed to enhance credit evaluation by mirroring real-world decision-making processes. The framework employs a layered architecture where specialized LLM-based agents collaboratively tackle sub-tasks. Additionally, we integrate contrastive learning for risk and reward assessment to optimize decision-making. We further present a signaling game theory perspective on hierarchical multi-agent systems, offering theoretical insights into their structure and interactions. Our paper also includes a detailed bias analysis in credit assessment, addressing fairness concerns. Experimental results demonstrate that MASCA outperforms baseline approaches, highlighting the effectiveness of hierarchical LLM-based multi-agent systems in financial applications, particularly in credit scoring.
arXiv:2607.05545v1 Announce Type: new Abstract: LLM conformity is often used to describe cases where a model changes a correct answer toward a peer or group response. We show that most of this apparent conformity survives even after the peer is removed. The reason is a confound: standard conformity prompts mix two cues at once, the presence of a speaker and the repeated wrong answer itself. Existing benchmarks vary these cues together, so they cannot tell how much of the revision actually depends on the speaker. We introduce a no-source condition: the same asserted answer with the explicit speaker removed. Across six open-weight LLMs and seven QA and reasoning datasets, this condition alone causes harmful revision in $66.5\%$ of initially correct cases, compared with $10.3\%$ under a plain re-ask. The effect also remains when the repeated answer is paraphrased and when answer options are hidden in an open-ended setting. Source framing mainly modulates this floor: expert-panel framing raises it, while minimal person labels do not reliably raise it. When models flip, they are usually confidently wrong, and simple recalibration does not recover the original answer. Source attribution still matters, but it should be measured as an increment above this speaker-free floor. The methodological lesson is that conformity benchmarks should first measure what remains after the speaker is removed; without this step, benchmarks may mistake repeated text for social influence.
arXiv:2607.05682v1 Announce Type: new Abstract: LLM systems for scientific discovery increasingly assist with ideation, literature synthesis, experiment planning, and report generation, but the first research question they propose can remain difficult to audit: it may sound plausible without exposing the mechanism, falsifier, or assumption that a scientist should inspect. We introduce FirstResearch, a first-principles research-question formation framework for scientific LLM agents whose core artifact is a structured Research Question Certificate. The certificate records primitive definitions, assumptions, a mechanism model, a tension or contradiction, a falsifiable hypothesis, a minimal decisive test, and a failure update rule, making the proposed question inspectable before downstream execution. On ten LLM-agent research topics, FirstResearch outperforms controlled prompt-level baselines inspired by AI co-scientist, Agent Laboratory, and AI Scientist-v2 under a primary DeepSeek-blind-judge protocol. A Gemini-2.5-Flash independent-judge rescore of the same 40 baseline packages preserves the system-level ranking, with FirstResearch scoring 4.86/5 versus 4.38/5 for the strongest baseline and Pearson agreement of 0.865 on average score. A one-repeat ablation checkpoint further suggests that the certificate-centered core is the strongest component: certificate-only scoring reaches 4.90/5 under DeepSeek and 4.88/5 under Gemini, while removing certificates drops below 1/5 under both judges. These results are preliminary and use LLM judges rather than human domain experts, but they support a narrow scientific-discovery claim: explicit derivation constraints are a promising mechanism for making LLM-generated scientific questions more auditable. Code, prompts, saved outputs, and reproduction scripts are available at https://github.com/louiswang524/FirstResearch.