Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

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

01.
arXiv (quant-ph) 2026-06-24

Entanglement in the Dicke subspace

arXiv:2602.15800v2 Announce Type: replace Abstract: We provide a complete mathematical theory for the entanglement of mixtures of Dicke states. These quantum states form an important subclass of bosonic states arising in the study of indistinguishable particles. We introduce a tensor-based parametrization where the diagonal entries of these states are encoded as a symmetric tensor, enabling a direct translation between entanglement properties and well-studied convex cones of tensors. Our results bridge multipartite entanglement theory with semialgebraic geometry and the theory of completely positive and copositive tensors. This dictionary maps separability to completely positive tensors, the PPT property to moment tensors, entanglement witnesses to copositive tensors, and decomposable witnesses to sum of squares tensors. Using this framework, we construct explicit PPT entangled states in three or more qutrits, disproving a recent conjecture. We establish that PPT entanglement exists for all multipartite systems with local dimension d >= 3 and n >= 3 parties. We also show that, for mixtures of Dicke states, the PPT condition with respect to the most balanced bipartition implies all other PPT conditions. We further connect bosonic extendibility of mixtures of Dicke states to the duals of known hierarchies for non-negative polynomials, such as the ones by Reznick and Polya. We thus provide semidefinite programming relaxations for separability and entanglement testing in the Dicke subspace.

03.
arXiv (CS.LG) 2026-06-11

Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering

arXiv:2606.12299v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models provide a natural language interface to robot control, but the mapping from language to behavior is often brittle and unintuitive: semantically similar instructions can induce drastically different behaviors, while some capabilities may not be elicitable through prompting alone. As a result, both human instructions and zero-shot language models can fail to reliably steer VLAs toward successful task execution. In this work, we propose a framework that interactively searches for language sequences that improve closed-loop VLA task performance, distills these sequences into a test-time language feedback policy (LFP), and learns an improvement head that predicts when language steering will improve performance. We conformalize this improvement head to prevent harmful steering interventions, where the LFP decreases task performance relative to the original instruction on out-of-distribution scenarios. Crucially, our approach operates on arbitrary frozen pre-trained VLAs, requiring neither access to the original training distribution nor fine-tuning of the underlying model. On seen environments, our conformalized LFP improves base VLA performance by 24.7% in simulation and 65.0% in hardware. On visual and semantic perturbations, our conformalized LFP has strong harmlessness guarantees, and produces recovery behaviors not observed with open-loop prompting.

04.
arXiv (quant-ph) 2026-06-11

Fun with Graph States: Nonlocal Bell Pairs and the Arf Invariant

arXiv:2606.06582v2 Announce Type: replace Abstract: We study inner products and partial amplitudes of graph states–a commonly employed class of quantum states, which are specified by graphs. We find that the magnitudes of these quantities are simply related to the rank of the adjacency matrix of the graph over F_2 while the phase is determined by the Arf invariant of its quadratic refinement. These facts motivate a nonlocal tensor factorization of the Hilbert space, with respect to which all graph states are products of Bell pairs with unentangled ancillae. These results may illuminate the quantum advantage in the framework of Measurement-Based Quantum Computation and suggest that graph states can be usefully visualized in the language of algebraic topology. In addition, we develop a specialized technique for computing expectation values of qubit-wise permutations in graph states, which is useful for calculating multi-invariants.

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

Cluster LOCO: Feature Importance For Interpreting Clusters

arXiv:2606.14592v1 Announce Type: cross Abstract: Clustering is widely used for exploratory analysis and scientific discovery, driving insights from market segmentation to biological data analysis, but its outputs can be difficult to interpret, audit, and reproduce as modern datasets become increasingly large and complex. Reliable use of clustering requires understanding which features drive the discovered structure, yet feature-level explanations for clustering remain scarce compared with methods in supervised learning. Furthermore, existing clustering feature importance scores are often tied to specific algorithms and data assumptions. To address these challenges, we propose Cluster LOCO (Leave-One-Covariate-Out), a family of model-agnostic feature importance scores for clustering. Cluster LOCO is built on feature occlusion and clustering generalizability, defined as whether cluster labels learned on one subset of the data can be accurately predicted on held-out samples. For any chosen clustering algorithm, Cluster LOCO quantifies a feature's importance by measuring how much its removal degrades generalizability. We first introduce Cluster LOCO-Split, which relies on data splitting, and then extend it to Cluster LOCO-MP, a minipatch ensemble-based version designed for large-scale data. Across synthetic simulations and an application to cell-type discovery in single-cell transcriptomics, we show that Cluster LOCO more reliably recovers informative features than existing clustering feature importance methods.

07.
arXiv (CS.CL) 2026-06-18

Simulating Hate Speech Cascades with Multi-LLM Agents: Empirical Grounding, Modeling Fidelity, and Intervention Strategies

Authors:

Faithful modeling of hateful content propagation on online platforms remains an open problem for moderation research. Classical cascade models that do not explicitly represent the profile, community, and content factors associated with hateful-content propagation may yield moderation strategies that behave less effectively when deployed in real-world scenarios. Multi-agent large language model (LLM) systems can, in principle, make each reshare decision depend on the user's profile, the surrounding community, and the post's content, but it remains unclear whether this added flexibility actually reproduces real hateful cascades more faithfully than classical baselines. We study three hateful Bluesky cascades and a size-matched benign control. In the empirical Bluesky data, we found that: 97.4–99.7\% of reposters take a hostile stance; toxicity-engagement homophily is higher on the diffusion tree than on the follower graph for hateful cascades; topology is star-like for the hateful cascades (most reposts come directly from the root) versus tree-like for the benign cascade (reposts propagate through multi-hop chains). In simulation, a multi-LLM-agent simulator reproduces the stance monoculture and the toxicity-delta direction. A structured ablation identifies agent heterogeneity as the leading fidelity factor, and amplifier targeting on dense networks yields 7.5–12.9\% reduction at 5.7\% benign collateral.

08.
arXiv (CS.CV) 2026-06-15

Naive Visual Memory is Not Enough: A Failure-Mode Study of GUI Agents

Graphical User Interface (GUI) agents are increasingly used to automate complex computer tasks across applications, websites, and operating systems. To improve their reliability, recent work has introduced experiential memory, where agents retrieve prior trajectories to guide decision-making in similar states. More recent approaches further extend this idea to visual memory by storing and retrieving screenshots from past interactions, providing agents with richer contextual information than text-only memories. However, the effect of visual memory in GUI agents remains insufficiently understood: it is unclear which failures visual memory mitigates, or which failures it exacerbates. To systematically analyze the effect of visual memory, we introduce a taxonomy of four GUI agent failures (i.e., cognitive failure, visual state misunderstanding, hidden operation blindness, and grounding error) that map to distinct stages of the perception-reasoning-action pipeline. We find that prepending full-image memory has a divergent effect on the failure distribution: it reduces state-level failures but worsens action-level ones, and increases hidden operation blindness and grounding error. Motivated by this finding, we propose Action-Grounded Visual Memory (AGMem), an action-grounded memory framework for GUI agents. The core idea of AGMem is to store image crops that capture the local GUI region closely related to a successful action or a recovery, rather than storing full screenshots. Experiments on OSWorld show that AGMem improves task success rates by 33.3 % over full-image memory. These results demonstrate that AGMem is an effective representation for visual memory in GUI agents.

09.
arXiv (CS.AI) 2026-06-12

SAIGuard: Communication-State Simulation for Proactive Defense of LLM Multi-Agent Systems

arXiv:2606.12474v1 Announce Type: cross Abstract: LLM-based multi-agent systems (MAS) solve complex tasks through inter-agent collaboration, but their communication-driven nature also allows security risks to spread across agents and trigger system-wide failures. Existing MAS defenses mainly follow a reactive paradigm after execution by detecting and isolating harmful agents, which may cause irreversible damage and degrade collaborative utility. To address this, we propose a proactive defense framework for MAS security, namely a Simulation-aware Interception Guard (SAIGuard). SAIGuard performs communication-state simulation over the MAS interaction graph, estimates the impact of incoming messages on local agent states and the global MAS state, and detects risky messages via reconstruction deviations from benign communication patterns. Instead of isolating agents, SAIGuard sanitizes or regenerates suspicious messages before it propagation into system. Experiments across diverse topologies and attack scenarios show that SAIGuard reduces attack success rates while maintaining MAS utility, outperforming reactive defenses.

10.
arXiv (CS.AI) 2026-06-16

Policy Regret for Embedding Model Routing: Contextual Bandits with Low-Rank Experts

arXiv:2606.14929v1 Announce Type: cross Abstract: Modern recommendation systems increasingly rely on dynamically routing diverse queries to multiple embedding models. Despite its practical significance, this problem remains poorly understood under realistic conditions like adversarial queries, bandit feedback, and limited observability of models. We formalize embedding model routing as an adversarial contextual linear bandit with low-rank experts, where contexts are queries, actions are items, and experts are the embedding models working on low-rank latent representation spaces. We first establish that standard regret notions suffer from structural misspecification or statistical intractability, and we identify a log-quadratic policy class that is expressive enough to capture query-dependent model routing, yet structured enough to allow efficient online learning. Second, we propose a policy gradient algorithm called Hypentropy Policy Gradient (HPG). It provably adapts to the unknown low-rank structure under incomplete information and attains $\tilde{\mathcal O}(s\sqrt{M T})$ linearized policy regret – where $s, M$, and $T$ are the intrinsic rank of the experts, the number of models, and the number of rounds – thus avoiding a curse of dimensionality. Finally, we also provide an computationally efficient and parameter-free implementation of HPG.

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

QoS-Aware Token Scheduling and Private Data Valuation for Multi-Modal Agentic Networks

arXiv:2606.15573v1 Announce Type: new Abstract: In agentic systems, human-generated data records anchor the value of AI services. Yet cloud compute pipelines centralize processing on remote servers. Data centralization reduces personal data sovereignty and may potentially degrade the quality of service (QoS). Meanwhile, user contributions are diverse in quantity and quality: decentralized records can be biased, noisy, and heterogeneously distributed. To address the data challenge, we study fair token allocation and private data valuation for decentralized and resource-constrained agentic systems. Our approach embeds multi-modal representations in a shared semantic space and releases differentially private (DP) prototypes to preserve utility while reducing semantic leakage. With the DP guarantee, we design a fair token allocation scheme that rewards effective contributions and remains robust to data heterogeneity and AI resource scarcity. Extensive simulations demonstrate improved contribution-based fairness and QoS compared to standard benchmarks. The improved resistance to image reconstruction attacks indicates enhanced privacy for multi-modal personal data.

12.
arXiv (CS.LG) 2026-06-16

Analytic Torsion and Spectral Gap Capture Persistent-Laplacian Performance

arXiv:2606.16990v1 Announce Type: new Abstract: While persistent Laplacians (PL) offer a richer geometric representation of data than persistent homology, utilizing their full eigenspectrum for learning tasks is often hampered by high dimensionality and the ``varying length'' problem across different filtration scales. We propose a compact spectral representation that distills the persistent Laplacian into three mathematically grounded invariants: Betti numbers, the spectral gap, and analytic torsion. Across benchmark datasets including MNIST, QM-3D, and SKEMPI WT, we demonstrate that this reduced feature space captures the essential predictive signal of the full spectrum, and in some cases outperforms it, while significantly reducing computational overhead and preventing the noise introduced by higher-frequency eigenvalues. Our results suggest that these invariants provide a principled, fixed-length interface between spectral geometry and topological learning.

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

X-MADAM-RAG: Diagnosing and Handling Chinese-English Evidence Conflict in Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) systems may receive evidence that is not merely noisy but mutually contradictory. This issue becomes particularly salient in multilingual settings, where retrieved Chinese and English evidence may support incompatible answer candidates. We study this problem through X-RAMDocs-ZHEN, a controlled Chinese-English benchmark derived from RAMDocs for diagnosing evidence conflict in RAG. The benchmark contains 300 examples across six balanced conditions, including monolingual support, bilingual agreement, reversed conflict directions, and conflict with optional noise. We further examine X-MADAM-RAG, an interpretable pipeline that decomposes evidence handling into per-document candidate extraction, visible-evidence repair, deterministic candidate grouping, and conflict-aware aggregation. On the original controlled benchmark with Qwen2.5-7B-Instruct, X-MADAM-RAG achieves 0.9667 strict accuracy and 0.9767 conflict-aware success, outperforming an evidence-normalized single-call baseline. However, a zero-call rule-only extractor reaches 1.0000 on the same benchmark, revealing strong template regularity. To probe this limitation, we construct a deterministic naturalized stress test that removes explicit answer templates while preserving candidate strings. On its 100-sample subset, rule-only extraction falls to 0.0000, but X-MADAM-RAG also drops to 0.3000 strict accuracy, below both naive and evidence-normalized baselines. A privileged oracle remains perfect, indicating that document-level extraction is the main bottleneck. These findings position X-RAMDocs-ZHEN and X-MADAM-RAG as diagnostic tools for controlled evidence conflict rather than as evidence of general hallucination detection or robustness to natural retrieval.

14.
arXiv (CS.LG) 2026-06-17

Can Vision Foundation Models Navigate? Zero-Shot Real-World Evaluation and Lessons Learned

arXiv:2603.25937v2 Announce Type: replace-cross Abstract: Visual Navigation Models (VNMs) promise generalizable, robot navigation by learning from large-scale visual demonstrations. Despite growing real-world deployment, existing evaluations rely almost exclusively on success rate, whether the robot reaches its goal, which conceals trajectory quality, collision behavior, and robustness to environmental change. We present a real-world evaluation of five state-of-the-art VNMs (GNM, ViNT, NoMaD, NaviBridger, and CrossFormer) across two robot platforms and five environments spanning indoor and outdoor settings. Beyond success rate, we combine path-based metrics with vision-based goal-recognition scores and assess robustness through controlled image perturbations (motion blur, sunflare). Our analysis uncovers three systematic limitations: (a) even architecturally sophisticated diffusion and transformer-based models exhibit frequent collisions, indicating limited geometric understanding; (b) models fail to discriminate between different locations that are perceptually similar, however some semantics differences are present, causing goal prediction errors in repetitive environments; and (c) performance degrades under distribution shift. We will publicly release our evaluation codebase and dataset to facilitate reproducible benchmarking of VNMs.

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

Towards Personalized Federated Learning for Dysarthric Speech Recognition

arXiv:2606.13253v1 Announce Type: cross Abstract: Speech recognition is challenging for dysarthric speakers. While federated learning (FL)-based ASR can be an effective tool for protecting privacy, it suffers from heterogeneity issues caused by speaker variability. Forcing all speakers to share the same model components can be suboptimal under such heterogeneity, making personalization a promising direction; however, related research on dysarthric speech remains limited. To this end, this paper explores two aggregation strategies to achieve personalization, including the parameter-based averaging strategy and the embedding-based averaging strategy. Experiments on UASpeech and TORGO show that the proposed methods outperform the baseline regularized FedAvg by statistically significant WER reductions of up to 0.99% absolute (3.15% relative) on UASpeech and 0.56% absolute (4.73% relative) on TORGO, respectively.

16.
arXiv (quant-ph) 2026-06-12

Driven-dissipative entanglement of distant giant atoms

arXiv:2606.13375v1 Announce Type: new Abstract: Quantum interconnects distribute entanglement via controlled light-matter interactions for quantum computing and sensing applications. Many entanglement generation schemes use coherent, reversible interactions that require precisely calibrated pulses to execute. In contrast, driven-dissipative protocols use a continuous-wave drive in the presence of correlated dissipation to stabilize entanglement in protected (dark) states. However, the same dissipation that generates the entanglement also limits its utility once the stabilization protocol ends. Here, we engineer a superconducting system of two giant artificial atoms coupled sequentially to a waveguide, with tunable individual and correlated dissipation enabled by interference between coupling points. Continuously driving the atoms through the waveguide exploits correlated dissipation to generate remote entanglement. We then tune the qubit frequencies in situ to suppress individual dissipation and thereby preserve the entanglement, achieving a Bell-state fidelity F = 0.89 +/- 0.02. This demonstration indicates that the driven dissipation of giant atoms is a viable approach for distributing entanglement across quantum networks.

17.
arXiv (quant-ph) 2026-06-24

Reachability and optimal-time certificates for quantum control

arXiv:2606.24645v1 Announce Type: new Abstract: Finite-time control is central to quantum technologies, yet rigorous limits on reachable targets and optimal control times remain largely unknown. We develop a framework for finite-time reachability and optimal-time certificates in constrained quantum control based on moment relaxations with implicitly time-dependent differential constraints. For fixed control horizons and control constraints, the method yields rigorous upper bounds on achievable terminal fidelities, lower bounds on the optimal control times required to reach them, and certificate gaps for benchmarking explicit control pulses. We demonstrate the versatility of our framework in three use cases: entangled-state preparation in two and three qubits, one-qubit gate synthesis across different control geometries, and excitation transfer in an $N$-qubit $XX$ chain. Our work establishes differential moment hierarchies as a practical tool for certifying reachability limits and optimal control times in quantum control, providing hardware-aware quantum speed limits while highlighting structure exploitation as a key ingredient for scalable certification.

18.
arXiv (CS.CV) 2026-06-24

Autonomous Video Generation with Counterfactual Controllability for Self-Evolving World Models

Existing literature claims that video generation essentially is world modelling. On the one hand, the claim is productive because it pushes generative AI beyond static images and toward temporally extended physical scenes. On the other hand, this claim dangerously relies on the belief that scaling visual prediction alone will automatically yield physical agents. We prefer a more accurate statement: video generation models learn a partial, implicit spatiotemporal world model, but not a fully grounded or controllable one. The reason is as follows: a model may generate a plausible video of a drone crossing a forest or a robot arm manipulating a cup, yet still fail to know which variables are controllable, which constraints belong to a particular body and which futures remain valid under intervention. The frontier in essence is not predictive realism alone, instead it emphasizes a self-evolving generative nature that requires the decisive criterion to be counterfactual controllability: the capability of asking what would happen under an action, to test whether the generated future can survive embodiment constraints and to feed the resulting action knowledge back into future imagination (generation). Therefore, in this paper we present a new perspective, i.e., autonomous video generation with counterfactual controllability is one promising way to realize self-evolving world models.

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

Closing the Calibration Gap in Semantic Caching

Semantic caching cuts LLM inference costs by serving a cached response to semantically similar queries. Standard practice evaluates these systems using PR-AUC, a metric that only measures how well scores rank and ignores whether they are usable at a fixed threshold. We show this mismatch leads to systematically poor deployment choices, as models with the highest PR-AUC are often the worst in operation. We introduce Precision-Cache Hit Ratio (P-CHR) AUC, a cache-aware metric that measures precision across cache utilization levels, and Calibration Retention Rate (CRR), which captures how much offline ranking quality survives at deployment. We decompose the operational gap between offline and deployed quality into a recoverable calibration component and an irreducible structural component fixed by the dataset's positive rate. Our experiments show that the calibration gap is governed by the training objective rather than data scale, and post-hoc calibration only partially closes it. Ultimately, model selection for semantic caching is a calibration problem, not a ranking one, and measuring it is the first step to closing the gap.

21.
arXiv (CS.CV) 2026-06-15

SA4Depth: Consistent Pose-Depth Scale Alignment for Self-Supervised Monocular Depth Estimation

Self-supervised depth estimation from monocular sequences relies on the joint learning of a depth and a pose network. Despite abundant research done to improve the depth network, efforts on the pose remain limited. In this context, even when depth is estimated up to scale, we highlight the importance of the alignment between the scene scales estimated by the pose and depth nets. Then, we introduce SA4Depth, an approach to improve this alignment and boost the depth predictions while keeping the inference time unchanged. Our proposed method uses the depth estimated during training to reproject learnable visual features across consecutive frames and refine the pose estimates by reducing feature alignment residuals. With our method, the estimated scene scales by the separate depth and pose networks are aligned, and the prediction scale consistency is improved across different sequences. Our differentiable refinement integrates seamlessly into existing self-supervised pipelines and substantially improves their depth estimates. We demonstrate this with extensive experiments both outdoors and indoors on KITTI, Cityscapes, and NYUv2. Additionally, results on KITTI Odometry confirm the effectiveness of our pose refinement. Our code is available at https://github.com/Runningchauncey/SA4Depth .

22.
arXiv (math.PR) 2026-06-12

Scaling limits of the single-curve interface and outermost loops in the planar random field Ising model

arXiv:2606.13147v1 Announce Type: new Abstract: We prove that the interface separating $+1$ and $-1$ spins in the near-critical planar random field Ising model (RFIM) with Dobrushin boundary conditions has a scaling limit, whose law is conformally covariant and almost surely absolutely continuous with respect to SLE$_3$. The limiting curve can be seen as a massive version of SLE$_3$ in the sense of Makarov and Smirnov, but in a random environment. We then show that the outermost spin loops of the near-critical planar RFIM with $+1$ boundary conditions have subsequential limits and that any of these limits is almost surely singular with respect to CLE$_3$. This dichotomy between absolute continuity of the single interface and singularity of the outermost loops reflects the fact that a single interface does not explore enough of the magnetization field of the near-critical RFIM to detect the singularity of this field with respect to the critical Ising magnetization field, whereas the outermost spin loops do.

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

Beyond Fully Random Masking: Attention-Guided Denoising and Optimization for Diffusion Language Models

Diffusion large language models (dLLMs) offer an efficient alternative to autoregressive models through parallel decoding, yet existing post-training methods largely rely on random masking strategies that overlook intrinsic token dependencies. In this work, we present an empirical analysis of attention in dLLMs and show that tokens attending more strongly to unmasked context exhibit greater generation stability and play a critical role in reasoning. Motivated by these findings, we propose AGDO, an attention-guided denoising and optimization framework that aligns both training and optimization with attention-derived dependencies. AGDO determines the denoising order based on attention structure and emphasizes attention-critical tokens during supervised fine-tuning and reinforcement learning. Experiments on mathematical and coding benchmarks demonstrate that AGDO consistently improves reasoning performance, outperforming state-of-the-art post-training methods for dLLMs.

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

Repeated Shared Access Enables Grokking, but Edit Propagation Depends on an Addressable Memory

Authors:

arXiv:2606.20737v2 Announce Type: replace Abstract: We study factual edit propagation in a controlled synthetic knowledge-graph QA setting using a 2x2 grid that crosses loop recurrence with shared-memory access: a dense transformer (Dense), a looped transformer (Loop), a dense backbone with shared memory (Dense+Mem), and a looped backbone with shared memory (loop-memory coupling, LMC). The two factors dissociate. For learning, both routes to repeated shared access – looped recomputation and repeated memory rereading – cross the out-of-distribution (OOD) grokking barrier that Dense fails, so repeated shared access is the behavioral regularity, not a specific architecture. For editing, the substrates split along a different axis: applying a single localized factual edit (conditioned on direct success) and measuring 2-hop propagation on a shared pre-edit-correct set, the edit propagates strongly in both memory-bearing cells (LMC 0.78-0.92, Dense+Mem 0.71-0.96) and only weakly in the memory-free ones (Loop 0.04-0.30, Dense 0.00-0.03). The split is along the memory axis, not the loop axis: every memory-bearing seed exceeds every memory-free seed, with no detectable difference between the two memory cells. Crucially Dense+Mem has no recurrence, so the propagating ingredient is an addressable site that an edit can write to and later computation rereads, not loop recomputation; Loop is at best a partial intermediate. The affordance survives coarsening the store (N=128 to N=13): propagation attenuates but the memory/no-memory split persists, so fine granularity buys precision rather than the affordance itself. These results dissociate learning competence from editing affordance – repeated shared access suffices to grok, but edit propagation depends on whether the substrate exposes an addressable memory that the forward computation can write to and later reread, an affordance that loop recurrence provides only partially.

25.
arXiv (CS.AI) 2026-06-12

Mental-R1: Aligning LLM Reasoning for Mental Health Assessment

arXiv:2606.13176v1 Announce Type: new Abstract: Mental health problems such as anxiety, depression, and suicide remain urgent global challenges, where timely and accurate assessment is critical for effective intervention. Recently, large language models have been explored for mental health assessment. However, existing general-purpose post-training methods do not align with the cognitive processes of human assessment, which may lead to unreliable reasoning outcomes. To bridge this gap, we propose Cognitive Relative Policy Optimization (CRPO), a reinforcement learning framework tailored for the mental health domain. CRPO extends group relative policy optimization by integrating stage-dependent uncertainty modeling into the policy optimization process. Specifically, we introduce a stage-wise entropy regularization mechanism that encourages broad exploration in early reasoning phases and progressively enforces confident decision-making in later stages, mimicking the human cognitive shift from uncertainty to certainty. In addition, inspired by cognitive appraisal theory, we formalize cognitive reasoning stages, thereby guiding theory-grounded interpretable inference. Experiments on 8 mental health datasets show that CRPO achieves an average improvement of 10.4 percentage points in weighted F1-score over the best reinforcement learning baseline. Furthermore, the CRPO-trained model Mental-R1 demonstrates clear advantages compared with existing large language models on reasoning-intensive cases, suggesting that CRPO enhances reasoning capabilities for mental health assessment.