Machine Learning
Local Inverse Geometry Can Be Amortized
Aaditya L. Kachhadiya
Nonlinear inverse problems often trade inexpensive but fragile first-order updates against
curvature-aware methods such as Gauss-Newton and Levenberg-Marquardt, which obtain stronger directions by repeatedly solving Jacobian-based linearized systems. We propose a learned alternative: amortize local inverse geometry into a reusable reverse operator. Our framework learns a bidirectional surrogate, Deceptron, and deploys it through D-IPG (Deceptron Inverse-Preconditioned Gradient), an iterative solver that pulls residual-corrected measurement-space proposals back to latent space. The key mechanism is a Jacobian Composition Penalty (JCP), which trains the reverse Jacobian to act as a local left inverse of the forward Jacobian; its runtime counterpart, RJCP, measures the same inverse-consistency error along optimization trajectories. We prove that D-IPG is first-order equivalent to damped Gauss-Newton under local pseudoinverse consistency, with deviation controlled by composition error and conditioning. Across seven PDE inverse-problem
benchmarks, D-IPG outperforms standard baselines, achieves 94.8% mean success across the six-problem reliability suite, and reaches comparable or better recovery quality at up to 77x lower inference-time solve cost on the main
benchmarks.
Read the full abstract ▾
Machine Learning
Ergodic Trajectory Design by Learned Pushforward Maps: Provable Coverage via Conditional Flow Matching
Ehsan Aghazadeh, Masoud Malekzadeh, Ahmad Ghasemi et al. (+1 more)
Designing continuous trajectories whose time-averaged occupancy provably matches a prescribed spatial density (the ergodic coverage problem) is central to
UAV-assisted data collection and sensing, robotic exploration, and mobile monitoring. For flying
agents in particular, this challenge is acute: trajectories must balance coverage fidelity against tight energy budgets, no-fly zones, and acceleration limits. Existing methods either re-optimize each trajectory online (with cost growing in the horizon and re-running for every target,
agent, and realization) or rely on bespoke analytical constructions that must be re-derived for each new constraint. We propose a epushforward framework that decouples ergodicity from density matching: an analytic latent trajectory provides exact uniform ergodicity on a simple annular domain, and a single map, learned offline by optimal-transport conditional flow matching, transports this latent occupancy onto the prescribed target density. The composed trajectory is then asymptotically ergodic with respect to the learned pushforward distribution, with deviation from the target controlled by the flow-matching training loss. Once trained for a given target density and constraint set, the map serves an unbounded number of trajectories and a multi-
agent fleet without per-
agent retraining, and many differentiable operational constraints (no-fly zones, acceleration ceilings, or
fairness penalties) enter as additive soft penalties in the training loss without re-deriving the design. We prove three results (an acceleration-energy bound, an $O(1/\sqrt{K})$ ergodic convergence rate in the number of trajectory cycles $K$, and an approximation-error bound) that combine into an end-to-end coverage bound estimable from CFM training diagnostics (certified given an architectural Lipschitz bound on $v_\theta$).
Read the full abstract ▾
All authors Ehsan Aghazadeh, Masoud Malekzadeh, Ahmad Ghasemi, Hossein Pishro-Nik
Machine Learning
Generative Support Realignment for Cross-Domain Offline Reinforcement Learning
Minung Kim, Jeongmo Kim, Gwanwoo Choi et al. (+1 more)
Cross-domain offline reinforcement learning learns a target policy from pre-collected source and target
datasets with different dynamics. When target data are scarce, effectively compensating for their limited coverage using source data remains challenging due to the discrepancy between domains. We propose Target-aligned Coverage Expansion (TCE), which leverages source states to generatively realign and expand the limited target support while controlling the generation error induced by this expansion. We further derive a performance gap bound that characterizes the interplay between generation error and source–target dynamics gap, providing theoretical guidance for effective coverage expansion and source utilization. Across diverse cross-domain environments, TCE consistently outperforms state-of-the-art baselines, while further analysis empirically supports the guidance provided by our theoretical findings.
Read the full abstract ▾
All authors Minung Kim, Jeongmo Kim, Gwanwoo Choi, Seungyul Han
Machine Learning
Large Language Models Lack Temporal Awareness of Medical Knowledge
Zihan Guan, Qiao Jin, Guangzhi Xiong et al. (+6 more)
The existing methods for evaluating the medical knowledge of Large
Language Models (LLMs) are largely based on atemporal examination-style
benchmarks, while in reality, medical knowledge is inherently dynamic and continuously evolves as new evidence emerges and treatments are approved. Consequently, evaluating medical knowledge without a temporal context may provide an incomplete assessment of whether LLMs can accurately reason about time-specific medical knowledge. Moreover, most medical data are historical, requiring the models not only to recall the correct knowledge, but also to know when that knowledge is correct. To bridge the gap, we built TempoMed-Bench, the first-of-its-kind
benchmark for evaluating the temporal awareness of the LLMs in the medical domain through evolving guideline knowledge. Based on the TempoMed-Bench, our evaluation analysis first reveals that LLMs lack temporal awareness in medical knowledge through the key findings: (1) model performance on up-to-date medical knowledge exhibits a gradual linear decline over time rather than a sharp knowledge-cutoff behavior, suggesting that parametric medical knowledge is not strictly bounded by knowledge cutoffs; (2) LLMs consistently struggle more with recalling outdated historical medical knowledge than with up-to-date recommendations: accuracy of historical knowledge is only 25.37%-53.89% of up-to-date knowledge, indicating potential knowledge forgetting effects during training; and (3) LLMs often exhibit temporally inconsistent behaviors, where predictions fluctuate irregularly across neighboring years. We also show that the temporal awareness problem is a challenge that cannot be easily solved when integrated with agentic search tools (-3.15%-14.14%). This work highlights an important yet underexplored challenge and motivates future research on developing LLMs that can better encode time-specific medical knowledge.
Read the full abstract ▾
All authors Zihan Guan, Qiao Jin, Guangzhi Xiong, Fangyuan Chen, Mengxuan Hu, Qingyu Chen, Yifan Peng, Zhiyong Lu, Anil Vullikanti
Machine Learning
What Information Matters? Graph Out-of-Distribution Detection via Tri-Component Information Decomposition
Danny Wang, Ruihong Qiu and Zi Huang
Graph neural networks are widely used for node classification, but they remain vulnerable to out-of-distribution (OOD) shifts in node features and graph structure. Prior work established that methods trained with standard supervised learning (SL) objectives tend to capture spurious signals from either features and/or structure, leaving the model fragile under distributional changes. To address this, we propose TIDE, a novel and effective Tri-Component Information Decomposition framework that explicitly decomposes information into feature-specific, structure-specific and joint components. TIDE aims to preserve only the label-relevant part of the joint information while filtering out spurious feature- and structure-specific information, thereby enhancing the separation between in-distribution (ID) and OOD nodes. Beyond the framework, we provide theoretical and empirical analyses showing that an information bottleneck objective is preferable to standard SL for graph OOD detection, with higher ID confidence and a greater
entropy gap between ID and OOD data. Extensive experiments across seven
datasets confirm the efficacy of TIDE, achieving up to a 34% improvement in FPR95 over strong baselines while maintaining competitive ID accuracy.
Read the full abstract ▾
Machine Learning
FeatCal: Feature Calibration for Post-Merging Models
Yanggan Gu, Shuo Cai, Zihao Wang et al. (+7 more)
Model merging combines task experts into one model and avoids joint training, retraining, or deploying many expert models, but the merged model often still underperforms task experts. We study this performance gap through feature drift, the difference between features produced by the merged model and by the expert on the same input. Our theory decomposes this drift into upstream propagation and local mismatch, tracks how it propagates and combines through later layers in forward order, and links final feature drift to output drift. This view motivates FeatCal, which uses a small calibration set to calibrate the merged model weights layer by layer in forward order, reducing feature drift while staying close to merged weights and preserving the benefits of model merging. FeatCal uses an efficient closed-form solution to update model weights, with no gradient descent, iterative optimization, or extra modules. On the main CLIP and GLUE
benchmarks, FeatCal beats Surgery and ProbSurgery, the closest post-merging calibration baselines: 85.5% vs. 77.0%/78.8% on CLIP-ViT-B/32 Task Arithmetic (TA) and 85.2% vs. 83.7%/82.2% on FLAN-T5-base GLUE. On CLIP-ViT-B/32, 8 examples per task reach 82.9%, and 256 examples per task take 53 seconds, about 4x faster than both baselines, showing better sample efficiency and lower calibration cost.
Read the full abstract ▾
All authors Yanggan Gu, Shuo Cai, Zihao Wang, Wenjun Wang, Yuanyi Wang, Pengkai Wang, Sirui Huang, Su Lu, Jianmin Wu, Hongxia Yang
Machine Learning
Understanding and Accelerating the Training of Masked Diffusion Language Models
Chunsan Hong, Sanghyun Lee, Chieh-Hsin Lai et al. (+5 more)
Masked
diffusion models (MDMs) have emerged as a promising alternative to autoregressive models (ARMs) for
language modeling. However, MDMs are known to learn substantially more slowly than ARMs, which may become problematic when scaling MDMs to larger models. Therefore, we ask the following question: how can we accelerate standard MDM training while maintaining its final performance? To this end, we first provide a detailed analysis of why MDM training is slow. We find that the main factor is the locality bias of
language: the predictive information for a
token is concentrated in nearby positions. We further investigate how this bias slows learning and suggest a simple yet effective remedy: bell-shaped time sampling as a training strategy. Notably, MDMs trained with our training recipe reach the same validation negative log-likelihood (NLL) up to $\sim4\times$ faster than standard training on One Billion Word
Benchmark (LM1B). We also show faster improvements in generative perplexity, zero-shot perplexity, and downstream task performance on various
benchmarks.
Read the full abstract ▾
All authors Chunsan Hong, Sanghyun Lee, Chieh-Hsin Lai, Satoshi Hayakawa, Yuhta Takida, Yuki Mitsufuji, Seungryong Kim, Jong Chul Ye
Machine Learning
Offline Two-Player Zero-Sum Markov Games with KL Regularization
Claire Chen, Yuheng Zhang, Xinyu Liu et al. (+3 more)
We study the problem of learning Nash
equilibria in offline two-player zero-sum Markov games. While existing approaches often rely on explicit pessimism to address distribution shift, we show that KL regularization alone suffices to stabilize learning and guarantee convergence. We first introduce Regularized Offline Sequential
Equilibrium (ROSE), a theoretical framework that achieves a fast $\widetilde{\mathcal{O}}(1/n)$ convergence rate under unilateral concentrability, improving over the standard $\widetilde{\mathcal{O}}(1/\sqrt{n})$ rates in unregularized settings. We then propose Sequential Offline Self-play Mirror Descent (SOS-MD), a practical model-free
algorithm based on least-squares value estimation and iterative self-play updates. We prove that the last iterate of SOS-MD attains the same $\widetilde{\mathcal{O}}(1/n)$ statistical rate up to a vanishing optimization error of order $\widetilde{\mathcal{O}}(1/\sqrt{T})$ in the number of self-play iterations $T$.
Read the full abstract ▾
All authors Claire Chen, Yuheng Zhang, Xinyu Liu, Zixuan Xie, Shuze Daniel Liu, Nan Jiang
Machine Learning
Rethinking Efficient Graph Coarsening via a Non-Selfishness Principle
Xu Bai, Bin Lu, Kun Zhang et al. (+4 more)
Graph coarsening is a graph dimensionality reduction technique that aims to construct a smaller and more tractable graph while preserving the essential structural and semantic properties of the original graph. However, most existing methods rely on pair-wise similarity matching, where each node independently searches for its best partner based on global information. This selfishness matching paradigm incurs substantial computational and
memory overhead. To address this problem, we shift to a non-selfishness principle that prioritizes the collective interference of neighborhood in coarsening, and propose an efficient method named NOPE, which achieves linear
memory consumption and near-linear computational complexity in the number of nodes. Furthermore, we derive a faster variant NOPE*, which reduces O(\delta \dot d) interference evaluation to O(d) based on the local isotropy assumption, and consequently alleviates the computational bottleneck for high-degree nodes. Experimental results show that NOPE* achieves 1.8-10\times speedup over NOPE and surpass almost all baselines with 1-3 orders of magnitude acceleration. Meanwhile, learning on coarsened graphs yields comparable performance to original graphs, and can even show superior performance over LLM-based graph
reasoning owing to compact graph information. The code can be available at https://github.com/dazonglian/NOPE-main.
Read the full abstract ▾
All authors Xu Bai, Bin Lu, Kun Zhang, Shengbo Chen, Xinbing Wang, Chenghu Zhou, Meng Jin
Machine Learning · Cross-listed from eess.IV
A General Bézier Tree Encoding Counterfactual Framework for Retinal-Vessel-Mediated Disease Analysis
Tan Su, Ethan Elio Meidinger, Lin Gu et al. (+1 more)
The geometry of the retinal vessel is a key biomarker of vascular diseases, yet clinical evidence remains primarily observational. Existing generative counterfactuals intervene only at the image-level disease label, failing to isolate explicit anatomical structure. To address this limitation, we propose the Bézier Tree Encoding Counterfactual Framework (BTECF). By abstracting vascular networks into interconnected cubic-Bézier segments, BTECF establishes a disease-agnostic representation in which structural
topology is explicitly preserved and atomically perturbable. Coupling this encoding with a
diffusion-based generator enables parameter-level do-interventions on explicit
geometric axes (e.g., tortuosity, caliber) while preserving background fundus textures. We validate BTECF on diabetic retinopathy, together with independent cohorts for ischemic stroke and Alzheimer's disease. Isolated counterfactual interventions produce dose-responsive shifts in classifier predictions; a matched pixel-drop control attenuates this response by an order of magnitude or more, ruling out out-of-distribution generation artifacts. By enforcing causal isolation between vessel
topology and pixel-level confounders, BTECF provides a unified generative paradigm for hypothesis verification across systemic diseases. To support reproducibility, the code will be publicly released upon acceptance.
Read the full abstract ▾
All authors Tan Su, Ethan Elio Meidinger, Lin Gu, Ruogu Fang
Machine Learning
JEDI: Joint Embedding Diffusion World Model for Online Model-Based Reinforcement Learning
Jing Yu Lim, Rushi Shah, Zarif Ikram et al. (+4 more)
Diffusion world models have recently become competitive for online model-based reinforcement learning, but current approaches expose a tension: pixel
diffusion is effective but computationally expensive while the latest latent
diffusion approach improves efficiency yet performs subpar. The latter also relies on separately trained latents rather than the end-to-end world-model objectives that have driven much of modern MBRL progress. In particular, JEPA-style predictive representation learning has emerged as an especially promising direction for world modeling and MBRL. Concurrently,
diffusion-style objectives have gained traction across multiple domains, with iterative refinement as a promising approach for
multimodal and
stochastic targets. Taken together, these trends motivate Joint
Embedding DIffusion (JEDI), the first online end-to-end latent
diffusion world model. JEDI learns its latent space directly from the
diffusion denoising loss with a JEPA framework, using denoising to learn and predict future latents rather than relying on reconstruction and pretrained models. We provide a theoretical motivation showing that conventional JEPA objectives induce a predictive information bottleneck, and that conditional
diffusion denoising admits a closely related predictive-compression decomposition. Empirically, JEDI is competitive on Atari100k and outperforms the baseline with seperately trained latents where directly comparable. Relative to the pixel
diffusion baseline, JEDI uses 43% less VRAM, over 3$\times$ faster world-model sampling, and 2.5$\times$ faster training. JEDI also exhibits a markedly different task-level performance profile from the pixel baseline, suggesting that end-to-end predictive latents change more than compute alone.
Read the full abstract ▾
All authors Jing Yu Lim, Rushi Shah, Zarif Ikram, Samson Yu, Haozhe Ma, Tze-Yun Leong, Dianbo Liu
Machine Learning · Cross-listed from q-bio.NC
Implicit Behavioral Decoding from Next-Step Spike Forecasts at Population Scale
John R. Minnick, Jesus Gonzalez-Ferrer, Kamran Hussain et al. (+6 more)
Closed-loop
brain-computer interfaces often require both a forecast of upcoming neural population activity and a readout of the animal's behavioral state. A single Mamba forecaster, trained only on next-step spike counts at Neuropixels scale, can deliver both in one forward pass. A lightweight per-session linear head reading the model's predicted rates decodes behavior better than the same linear classifier reading the raw spike counts, under matched temporal context. We test on the Steinmetz visual-discrimination
benchmark, which spans 39 sessions, roughly 27,000
neurons, and 1,994 held-out trials. Across three training seeds, Mamba's predicted rates decode mouse choice at 75.7$\pm$0.2% trial vote, roughly 2.3 times chance level, and stimulus side at 66.1$\pm$0.6%, about twice chance. Compared to a matched 500 ms-context linear decoder on the raw spike counts, Mamba wins at trial vote by 4-6 pp on response and 4-6 pp on stimulus side. A session-start calibration block of about 100-150 trials brings the readout within 1-2 pp of asymptote, and the full pipeline fits inside the 50 ms bin budget on workstation-class GPUs typical of tethered chronic Neuropixels recordings.
Read the full abstract ▾
All authors John R. Minnick, Jesus Gonzalez-Ferrer, Kamran Hussain, Jinghui Geng, Ash Robbins, Mohammed A. Mostajo-Radji, David Haussler, Jason Eshraghian, Mircea Teodorescu
Machine Learning
DRIFT: A Benchmark for Task-Free Continual Graph Learning with Continuous Distribution Shifts
Guiquan Sun, Xikun Zhang, Jingchao Ni et al. (+1 more)
Continual graph learning (CGL) aims to learn from dynamically evolving graphs while mitigating catastrophic forgetting. Existing CGL approaches typically adopt a task-based formulation, where the data stream is partitioned into a sequence of discrete tasks with pre-defined boundaries. However, such assumptions rarely hold in real-world environments, where data distributions evolve continuously and task identity is often unavailable. To better reflect realistic non-stationary environments, we revisit continual graph learning from a task-free perspective. We propose a unified formulation that models the data stream as a time-varying mixture of latent task distributions, enabling continuous modeling of distribution drift. Based on this formulation, we construct DRIFT, a
benchmark that spans a spectrum of transition dynamics ranging from hard task switches to smooth distributional drift through a Gaussian parameterization. We evaluate representative continual learning methods under this task-free setting and observe substantial performance degradation compared to traditional task-based protocols. Our findings indicate that many existing approaches implicitly rely on task boundary information and struggle under realistic task-free graph streams. This work highlights the importance of studying continual graph learning under realistic non-stationary conditions and provides a
benchmark for future research in this direction. Our code is available at https://github.com/UConn-DSIS/DRIFT.
Read the full abstract ▾
All authors Guiquan Sun, Xikun Zhang, Jingchao Ni, Dongjin Song
Machine Learning
Frequency Bias and OOD Generalization in Neural Operators under a Variable-Coefficient Wave Equation
Runlong Xie and An Luo
Neural operators learn to map initial conditions to the terminal solution of partial differential equations (PDEs), providing a surrogate for the full operator mapping. This enables rapid prediction across different input configurations. While recent neural operator architectures have demonstrated strong performance on diverse PDE tasks, their behavior under structured distribution shifts remains insufficiently understood. To investigate this, we study operator learning in a wave propagation setting governed by a one-dimensional variable-coefficient wave equation, using two representative architectures, the
Fourier Neural Operator (FNO) and the Deep Operator Network (DeepONet). To examine their generalization under distribution shifts, we consider structured out-of-distribution (OOD) settings that independently vary input frequency and coefficient smoothness. The results show that under smoothness shifts, both models maintain stable performance, with FNO achieving lower error. In contrast, under frequency shifts, FNO exhibits a sharp increase in error under unseen high-frequency inputs, whereas DeepONet shows milder degradation despite higher overall error. Our analysis reveals that these differences arise from how each architecture represents and responds to variations in frequency structure. Together, these findings highlight a fundamental gap between strong in-distribution performance and generalization under distribution shifts in operator learning, underscoring the role of architectural representation bias in developing more reliable neural operators for physics-based PDE simulations beyond the training distribution.
Read the full abstract ▾
Machine Learning
F-GRPO: Factorized Group-Relative Policy Optimization for Unified Candidate Generation and Ranking
Rohan Surana, Gagan Mundada, Junda Wu et al. (+9 more)
Traditional
retrieval pipelines optimize utility through stages of candidate
retrieval and reranking, where ranking operates over a predefined candidate set. Large
Language Models (LLMs) broaden this into a generative process: given a candidate pool, an LLM can generate a subset and order it within a single autoregressive pass. However, this flexibility introduces a new optimization challenge: the model must search a combinatorial output space while receiving utility feedback only after the full ranked list is generated. Because this feedback is defined over the completed sequence, it cannot distinguish whether a poor result arises from failing to generate a relevant subset or from failing to rank that subset correctly. This credit assignment gap makes end-to-end optimization unstable and sample-inefficient. Existing systems often address this by separating candidate generation from ranking. However, such decoupling remains misaligned with downstream utility because ranking is limited by the candidate set it receives. To bridge this gap, we propose a unified framework that performs both within a single autoregressive rollout and optimizes them end-to-end via factorized group-relative policy optimization (F-GRPO). Our framework factorizes the policy into candidate generation and ranking while sharing a single LLM backbone, and jointly trains them with an order-invariant coverage reward and a position-aware utility reward. To address the resulting phase-specific credit assignment problem, we use separate group-relative advantages for generation and ranking within a two-phase sequence-level objective. Across sequential recommendation and multi-hop question answering
benchmarks, F-GRPO improves top-ranked performance over GRPO and decoupled baselines, outperforms supervised alternatives, and remains competitive with strong zero-shot rerankers, with no architectural changes at inference time.
Read the full abstract ▾
All authors Rohan Surana, Gagan Mundada, Junda Wu, Xintong Li, Yizhu Jiao, Bowen Jin, Sizhe Zhou, Tong Yu, Ritwik Sinha, Jiawei Han, Jingbo Shang, Julian McAuley
Machine Learning
DP-Muon: Differentially Private Optimization via Matrix-Orthogonalized Momentum
Jihwan Kim and Chenglin Fan
We study differentially private optimization with
matrix-orthogonalized momentum. DP-
Muon uses conventional global per-example clipping and one Gaussian gradient release per step;
matrix updates and auxiliary updates are post-processing. Our main contribution concerns the additional mean distortion created when fresh Gaussian noise passes through a nonlinear
matrix map. Conditioning on the actual adaptive history immediately before the current noise yields an exact Gaussian heat identity. For a smooth Newton-Schulz map, first-order DP-MuonBC reduces this conditional output bias from second to fourth order in the fresh noise scale, and an arbitrary-order extension has bias of order $2K+2$. We prove
matrix-block stationarity bounds under global clipping, retain finite-step orthogonalization error explicitly, and give an exact criterion for improvement of the resulting upper bound. A separate inequality exposes the effect of auxiliary Adam updates. GPT-2 experiments on E2E at four
privacy targets favor the reported
Muon configurations over Adam baselines in test NLL.
Read the full abstract ▾
Machine Learning · Cross-listed from q-bio.NC
SpikeProphecy: A Large-Scale Benchmark for Autoregressive Neural Population Forecasting
John R. Minnick, Jinghui Geng, Kamran Hussain et al. (+6 more)
Neural population models, which predict the joint firing of many simultaneously recorded
neurons forward in time, are typically evaluated by a single aggregate Pearson correlation $r$ between predicted and actual spike counts, a number that masks critical structure. We argue that how we evaluate spike
forecasting matters as much as what we build, and introduce SpikeProphecy, the first large-scale
benchmark for causal, autoregressive spike-count
forecasting on real electrophysiology recordings. Our core contribution is a population metric decomposition that separates aggregate performance into temporal fidelity, spatial pattern accuracy, and magnitude-invariant
alignment. The decomposition surfaces aspects of the underlying data that an aggregate scalar collapses together. We apply the protocol to 105 Neuropixels sessions (Steinmetz 2019 + IBL Repeated Site; ~89,800
neurons) with seven architecture baselines spanning four structural families: four SSMs (three diagonal and one non-diagonal), a
Transformer, an LSTM, and a spiking network. The decomposition surfaces a
brain-region predictability ranking that reproduces across all seven baselines and survives ANCOVA correction for firing-statistics constraints (region $\Delta R^2 = 0.018$ above the firing-statistics
covariates). It also exposes a sub-Poisson evaluation floor where rigorous metrics combine with genuine biophysical constraints on regular spike trains, and yields a negative result on KL-on-output-rates
distillation for ANN-to-SNN transfer in this Poisson count domain.
Read the full abstract ▾
All authors John R. Minnick, Jinghui Geng, Kamran Hussain, Jesus Gonzalez-Ferrer, Ash Robbins, Mohammed A. Mostajo-Radji, David Haussler, Jason K. Eshraghian, Mircea Teodorescu
Machine Learning
Not Just RLHF: Why Alignment Alone Won't Fix Multi-Agent Sycophancy
Adarsh Kumarappan and Ananya Mujoo
LLM-based multi-
agent pipelines flip from correct to incorrect answers under simulated peer disagreement at rates we term yield, a vulnerability widely attributed to RLHF-induced sycophancy. We test this attribution across four model families and find it largely wrong: pretrained base models exhibit the same substitution pattern as their Instruct variants, averaging higher yield than Instruct. Using activation patching, we localize the corruption to a narrow mid-layer window where
attention carries the causal weight and MLP contribution is negligible; patching above this window restores 96% of the clean-to-pressured P(correct) gap. The attack surface decomposes into two independent factors (channel framing and consensus strength) whose interaction produces a 47.5 percentage-point yield gap at majority consensus, preserved across jury sizes $N \in \{4, 5, 6\}$. Two converging activation-space interventions show that pressure suppresses clean-
reasoning features rather than activating a new sycophancy
circuit. A single correctly-arguing dissenter reduces yield by 54-73 percentage points across all framings tested, whereas the strongest
prompt-level defense fails on attack variants outside its design surface. Mitigations should target the mechanism, structured dissent at the pipeline level, rather than
prompt-level defenses.
Read the full abstract ▾
Machine Learning
Decision Tree Learning on Product Spaces
Arshia Soltani Moakhar, Faraz Ghahremani, Kiarash Banihashem et al. (+1 more)
Decision tree learning has long been a central topic in theoretical computer science, driven by its practical importance. A fundamental and widely used method for decision tree construction is the top-down greedy heuristic, which recursively splits on the most influential variable. Despite its empirical success, theoretical analysis of this heuristic has been limited. A recent breakthrough by Blanc et al. (ITCS, 2020) provided the first rigorous theoretical guarantees for the greedy approach, but only under the uniform distribution. We extend this analysis to the more general and practically relevant setting of arbitrary product distributions. Our main result shows that for any function $f$ computable by an optimal decision tree of size $s$, maximum depth $D_{\text{opt}}$, and average depth $\Delta_{\text{opt}}$, the greedy heuristic constructs an $\epsilon$-approximating tree whose size grows at most with $\exp(\Delta_{\text{opt}} D_{\text{opt}} \log(e/\epsilon))$. In the special case where the optimal tree is a full binary tree, this bound improves upon the bound of Blanc et al. and holds under a strictly broader class of distributions. Moreover, we present an
algorithm based on the top-down greedy heuristic that is entirely parameter-free – it requires no prior knowledge of the optimal tree's size or depth – offering a practical advantage over Blanc et al.'s method.
Read the full abstract ▾
All authors Arshia Soltani Moakhar, Faraz Ghahremani, Kiarash Banihashem, MohammadTaghi Hajiaghayi
Machine Learning · Cross-listed from cs.SE
Protocol-Driven Development: Governing Generated Software Through Invariants and Continuous Evidence
Jun He and Deying Yu
Automated program synthesis lowers the cost of producing implementations but introduces a harder governance problem: determining which generated artifacts are admissible. Natural-
language specifications are ambiguous, and example-based tests sample only part of the behavioral space. Used alone, neither provides a sufficient control boundary. We introduce Protocol-Driven Development (PDD), where the primary software artifact is a machine-enforceable protocol rather than code. We define a protocol as the triplet P = (S, B, O), specifying structural, behavioral, and operational invariants. Their conjunction defines the admissible implementation space of a software component. Under PDD, implementations are replaceable realizations discovered through constrained search. An implementation is admitted only if it satisfies the protocol and produces a verifiable Evidence Chain of compliance. Admission is grounded in protocol satisfaction and recorded evidence rather than trust in the generator. For deployed systems, we extend the Evidence Chain into a Dynamic Evidence Ledger. Runtime verifiers append signed observations, invariant checks, and violations to the ledger, allowing monitorable obligations to be continuously attested. This connects live failures back to the generation loop without granting the generator runtime authority. Combining formal methods, property testing, runtime verification, policy-as-code, and software provenance, PDD defines a governance model for automated software engineering. Its organizing principle is that code is transient, while the protocol carries durable authority.
Read the full abstract ▾
Machine Learning
CoRe-Gen: Robust Spectrum-to-Structure Generation under Imperfect Fingerprint Conditions
Tianbo Liu, Chixiang Lu, Jing Hao et al. (+4 more)
Molecular structure elucidation from tandem mass spectra (MS/MS) remains challenging, particularly for de novo generation beyond database coverage. A common approach decomposes the task into spectrum-to-fingerprint prediction followed by fingerprint-to-structure decoding, enabling the use of large-scale molecular corpora. However, at deployment, the decoder relies on predicted rather than oracle fingerprints, introducing structured errors that propagate into generation. Moreover, common autoregressive decoders threshold fingerprint probabilities and serialize active bits as discrete input
tokens, blocking molecular generation gradients from the spectrum encoder. We present CoRe-Gen, which improves the intermediate condition through synthetic-spectrum pretraining, matches deployment-time noise through frequency-aware corruption, and introduces a differentiable fingerprint-to-
memory bridge for joint encoder–decoder finetuning. The bridge preserves soft bit confidences and conditions a structure-aware autoregressive decoder without discrete fingerprint tokenization. CoRe-Gen achieves 21.49%/35.42% Top-1/Top-10 exact-match accuracy on NPLIB1 and, under encoder-aligned evaluation, 20.02%/22.45% on MassSpecGym, while retaining efficient autoregressive inference.
Read the full abstract ▾
All authors Tianbo Liu, Chixiang Lu, Jing Hao, Hengyu Zhang, Lifei Wang, Haibo Jiang, Xiaojuan Qi
Machine Learning
Revisiting Reinforcement Learning with Verifiable Rewards from a Contrastive Perspective
Feng Zhang, Xinhong Ma, Ziqiang Dong et al. (+5 more)
Group Relative Policy Optimization (GRPO) is one of the most widely adopted RLVR
algorithms for post-training large
language models on
reasoning tasks. We first show that GRPO admits an equivalent discriminative reformulation, in which policy optimization maximizes the expected score gap between verified positive and negative rollouts. This reformulation reveals two objective-level limitations: likelihood-misaligned surrogate scores, in which clipped ratio-based scores are optimized rather than the sequence likelihoods that govern generation, and score-insensitive credit assignment, in which rollout-level credit does not reflect the current score gaps between positive and negative rollouts. To address these limitations, we propose ConSPO, a Contrastive Sequence-level Policy Optimization method that uses length-normalized sequence log-probabilities as rollout scores and contrasts verified positive rollouts against negative distractors within the same group. ConSPO optimizes a group-wise InfoNCE-style objective to adaptively strengthen updates for poorly separated positives and high-scoring negatives, together with a curriculum-scheduled margin that preserves separation pressure as training progresses. Experiments across diverse settings show that ConSPO outperforms strong baselines on challenging
reasoning benchmarks.
Read the full abstract ▾
All authors Feng Zhang, Xinhong Ma, Ziqiang Dong, Xi Leng, Jianfei Zhao, Xin Sun, Yang Yang, Guanjun Jiang
Machine Learning
Controlling Logical Collapse in LLMs via Algebraic Ontology Projection over F2
Hisashi Miyashita and Mgnite Inc
Do large
language models internally encode ontological relations in a formally verifiable algebraic structure? We introduce Algebraic Ontology Projection (AOP), which projects LLM hidden states into the Galois Field F2 under Liskov Substitution Principle constraints, using only 42 relational pairs as algebraic keys. AOP achieves up to 93.33% zero-shot inclusion accuracy on unseen concept pairs (Gemma-2 Instruct with optimized
prompt), with consistent 86.67% accuracy observed across multiple model families – with no model tuning, but through
prompt alone. This algebraic structure is strongly layer-dependent. We introduce Semantic Crystallisation (SC), a metric that quantifies F2 constraint satisfaction relative to a random baseline and predicts zero-shot accuracy without held-out data. System
prompts act as algebraic boundary conditions: only their combination with instruction tuning prevents Late-layer Collapse – a systematic degradation of logical consistency in the final layers, observed in 7 of 10 conditions. These findings reframe forward computation as an iterative process of algebraic organisation, and open a path toward LLMs whose logical structure is not merely approximated, but formally accessible.
Read the full abstract ▾
Machine Learning
U-HNO: A U-shaped Hybrid Neural Operator with Sparse-Point Adaptive Routing for Non-stationary PDE Dynamics
Yingzhe Ma, Xiao Yang, Yuxin Xie et al. (+2 more)
Solutions to many partial differential equations (PDEs) display coexisting smooth global transport and localized sharp features within a single trajectory: shock fronts, thin interfaces, and concentrated high-frequency content sit on top of slowly varying backgrounds. This poses a challenge for neural operators:
Fourier-based architectures mix nonlocal interactions efficiently but tend to under-resolve localized non-smooth features, whereas spatially local architectures recover fine detail at the cost of long-range propagation and rollout stability. Existing hybrid operators paper over this tension with a fixed, spatially uniform fusion that forces the same trade-off everywhere. We propose U-HNO, a U-shaped hybrid neural operator whose central design is Sparse-Point Adaptive Routing (SPAR): at every spatial location, a per-pixel hard mask selects whether the global
Fourier branch or the local multi-scale Gaussian branch should dominate, and the sparsity ratio is a function of the local contrast of the routing signal, so smooth and shock-aligned regions receive different mixtures of global and local computation. SPAR is embedded in a hierarchical encoder-bottleneck-decoder backbone with skip connections so that the dual branches and the gate operate at every resolution. Training combines pointwise supervision with a finite-difference H^1 gradient term and a band-wise spectral consistency regularizer. Across
benchmarks spanning 1D Burgers, Kuramoto-Sivashinsky, KdV, 2D advection, Allen-Cahn, Navier-Stokes, Darcy flow, and 3D transonic compressible Navier-Stokes from PDEBench, U-HNO achieves state-of-the-art rollout accuracy on the majority of tasks in both relative L^2 and H^1 metrics, with the largest gains on problems dominated by sharp localized features. Ablations show that removing any single component substantially degrades rollout error.
Read the full abstract ▾
All authors Yingzhe Ma, Xiao Yang, Yuxin Xie, Zihan Xiong, Jinliang Liu
Machine Learning · Cross-listed from cs.CV
Reducing Bias and Variance: Generative Semantic Guidance and Bi-Layer Ensemble for Image Clustering
Feijiang Li, Zhenxiong Li, Jieting Wang et al. (+3 more)
Image clustering aims to partition unlabeled image
datasets into distinct groups. A core aspect of this task is constructing and leveraging prior knowledge to guide the clustering process. Recent approaches introduce semantic descriptions as prior information, most of which typically relying on matching-based techniques with predefined vocabularies. However, the limited matching space restricts their adaptability to downstream clustering tasks. Moreover, these methods primarily focus on reducing bias to improve performance, frequently overlooking the importance of variance reduction. To address these limitations, we propose GSEC (Image Clustering based on Generative Semantic Guidance and Bi-Layer Ensemble), a framework designed to reduce bias through generative semantic guidance and mitigate variance via ensemble learning. Our method employs
Multimodal Large
Language Models to generate semantic descriptions and derive image
embeddings via weighted averaging. Additionally, a bi-layer ensemble strategy integrates cross-modal information through BatchEnsemble in the inner layer and aligns outputs via an
alignment mechanism in the outer layer. Comparative experiments demonstrate that GSEC outperforms 18 state-of-the-art methods across six
benchmark datasets, while further analysis confirms its effectiveness in simultaneously reducing both bias and variance. The code is available at https://github.com/2017LI/GSEC.git.
Read the full abstract ▾
All authors Feijiang Li, Zhenxiong Li, Jieting Wang, Zizheng Jiu, Saixiong Liu, Liang Du
Machine Learning · Cross-listed from stat.ML
Coreset-Induced Conditional Velocity Flow Matching
Zihua She, Jianxi Su and Xiao Wang
We propose Coreset-Induced Conditional Velocity Flow Matching (CCVFM), a generative model that augments hierarchical rectified flow with a data-informed source distribution. Standard flow matching regresses only the conditional mean of the velocity; hierarchical flow matching instead models the full conditional velocity law with a second, inner flow in velocity space, but this inner flow is asked to transport isotropic Gaussian noise to a
multimodal target velocity distribution from scratch. Our key observation is that this inner source can be replaced by a closed-form surrogate built from a coreset of the target. CCVFM first compresses the target into weighted atoms using an entropic Sinkhorn coreset and lifts them to a Gaussian mixture. The induced conditional velocity law is then a closed-form Gaussian mixture that can be sampled without a learned neural sampler. A lightweight correction flow, trained from this surrogate source, then refines the remaining surrogate-to-target residual rather than learning an entire noise-to-data map. We prove that the surrogate transport cost equals the target-surrogate Wasserstein gap under an explicit compression assumption, whereas the noise-source analogue has a dimension-scale lower bound. We further characterize the conditional second moment of the training target when the source velocity is drawn from the surrogate law independently of the target velocity, and show that the part of it that depends on the source is small when the surrogate conditional law is close to the true conditional velocity law in mean and covariance. Empirically, on MNIST, CIFAR-10, ImageNet-32, and CelebA-HQ, the proposed method reaches competitive few-step generation under matched architectures. Code is available at https://github.com/ziwaa-se/CCVFM.
Read the full abstract ▾
Machine Learning · Cross-listed from stat.ML
When Should an AI Workflow Release? Always-Valid Inference for Black-Box Generate-Verify Systems
Young Hyun Cho and Will Wei Sun
LLM-enabled AI workflows increasingly produce outputs through iterative generate-evaluate-revise loops. Each iteration can improve the candidate, but it also creates a release decision: when to stop and output the current result? This raises a statistical challenge because deployment-time evaluator scores are adaptively generated and repeatedly monitored, yet the likelihood models or exchangeability assumptions typically used for calibration are unavailable. We propose an always-valid release wrapper for existing generator-evaluator pipelines. The wrapper builds a hard-negative reference pool of high-scoring failures, calibrates deployment-time evaluator scores against this pool, and accumulates the resulting evidence with an e-process. This separates two roles: the reference pool turns black-box scores into conservative evidence, while the e-process provides validity under optional stopping. In theory, we show that a conservative reference pool yields finite-sample control of the probability of releasing on infeasible tasks, that is, tasks for which the given workflow is not capable of producing a reliable solution. We also characterize conditions under which the same conservative rule still achieves nontrivial release on feasible tasks. In an MBPP+ coding-
agent case study, the wrapper reduces premature incorrect release relative to baseline stopping rules while still releasing on tasks for which the workflow repeatedly accumulates moderate supporting evidence.
Read the full abstract ▾
Machine Learning
Separating Shortcut Transition from Cross-Family OOD Failure in a Minimal Model
Hongmin Li
Shortcut features are often invoked to explain out-of-distribution (OOD) failure, but training correlation, learned shortcut use, and test-time failure need not coincide. We study a minimal binary model with one invariant coordinate and one family-dependent shortcut coordinate. In the deterministic regime, positive average shortcut correlation pulls logistic ERM toward positive shortcut weight, but ridge regularization keeps the classifier invariant-dominated and prevents deterministic OOD failure. When the invariant coordinate is noisy, ridge-logistic ERM switches to the shortcut rule once the training shortcut signal exceeds the invariant signal. Whether that transition causes failure depends on the held-out family: weaker shortcut correlation yields positive excess risk, and sign-flipped families yield above-chance error. Synthetic checks match these analytic regimes and show that the same training-side transition can have different held-out consequences. The model separates shortcut attraction, shortcut-rule transition, and cross-family OOD failure.
Read the full abstract ▾
Machine Learning
From Instance Selection to Fixed-Pool Data Recipe Search for Supervised Fine-Tuning
Haodong Wu, Jiahao Zhang, Lijie Hu et al. (+1 more)
Supervised fine-tuning (SFT) data selection is commonly formulated as instance ranking: score each example and retain a top-$k$ subset. However, effective SFT training subsets are often produced through ordered curation recipes, where filtering, mixing, and deduplication operators jointly shape the final data distribution. We formulate this problem as fixed-pool data recipe search: given a raw instruction pool and a library of grounded operators, the goal is to discover an executable recipe that constructs a high-quality selected subset under a limited budget of full SFT evaluations, without generating, rewriting, or augmenting training samples. We introduce AutoSelection, a two-layer solver that decouples fixed-pool materialization based on cached task-, data-, and model-side signals from expensive full evaluation, using warmup probes, realized subset states, local recipe edits, Gaussian-process-assisted ranking, and stagnation-triggered reseeding. Experiments on a 90K instruction pool show that AutoSelection achieves the strongest in-distribution
reasoning average across three base models, outperforming full-data training, random recipe search, random top-$k$, and single-operator selectors. Additional Out-of-distribution graph-
reasoning results, search-stability analyses, structural ablations, and 1.5B-to-7B transfer checks further show that recipe structure matters beyond individual selection operators. Code is available at https://github.com/w253/AutoSelection.
Read the full abstract ▾
All authors Haodong Wu, Jiahao Zhang, Lijie Hu, Yongqi Zhang
Machine Learning
Reinforced Collaboration in Multi-Agent Flow Networks
Zheng Wang, Yuang Liu and Yangkai Ding
Multi-
agent systems provide a powerful way to extend large
language models (LLMs) by decomposing a complex task into specialized subtasks handled by different
agents. However, their performance is often hindered by error propagation, arising from suboptimal workflow design or inaccurate
agent outputs, which can propagate through the
agent collaboration process and degrade final results. To address the challenges, we present MANGO (Multi-
Agent Network Gradient Optimization), a data-driven framework that organizes and refines
agent collaboration via a flow network constructed from past successful workflows. MANGO integrates reinforcement learning and textual gradients to jointly optimize workflow paths and
agent behaviors, while a skipping mechanism prevents redundant updates to well-optimized
agents for improving efficiency. Extensive experiments on seven
benchmarks show that MANGO achieves up to 12.8% performance improvement over state-of-the-art baselines, enhances efficiency by 47.4%, and generalizes effectively to unseen domains. Our code and
datasets are publicly available at https://github.com/openJiuwen-ai/
agent-store/tree/main/community/mango.
Read the full abstract ▾