arXiv:2604.26904v3 Announce Type: replace
Abstract: Claw-style environments support multi-step workflows over local files, tools, and persistent workspace states. However, scalable development around these environments remains constrained by the absence of a systematic framework, especially one for synthesizing verifiable training data and integrating it with agent training and diagnostic evaluation. To address this challenge, we present ClawGym, a scalable framework that supports the full lifecycle of Claw-style personal agent development. Concretely, we construct ClawGym-SynData, a diverse dataset of 13.5K filtered tasks synthesized from persona-driven intents and skill-grounded operations, paired with realistic mock workspaces and hybrid verification mechanisms. We then train a family of capable Claw-style models, termed ClawGym-Agents, through supervised fine-tuning on black-box rollout trajectories, and further explore reinforcement learning via a lightweight pipeline that parallelizes rollouts across per-task sandboxes. To support reliable evaluation, we further construct ClawGym-Bench, a benchmark of 200 instances calibrated through automated filtering and human-LLM review. Relevant resources have been released at https://github.com/ClawGym.
Science Journals
arXiv:2601.20888v3 Announce Type: replace-cross
Abstract: We study sampling from posterior distributions in Bayesian linear inverse problems where $A$, the parameters to observables operator, is computationally expensive. In many applications, $A$ can be factored in a manner that facilitates the construction of a cost-effective approximation $\tilde{A}$. In this framework, we introduce Latent-IMH, a sampling method based on the Metropolis-Hastings independence (IMH) sampler. Latent-IMH first generates intermediate latent variables using the approximate $\tilde{A}$, and then refines them using the exact $A$. Its primary benefit is that it shifts the computational cost to an offline phase. We theoretically analyze the performance of Latent-IMH using KL divergence and mixing time bounds. Using numerical experiments on several model problems, we show that, under reasonable assumptions, it outperforms state-of-the-art methods such as the No-U-Turn sampler (NUTS) in computational efficiency. In some cases, Latent-IMH can be orders of magnitude faster than existing schemes.
arXiv:2605.17716v1 Announce Type: new
Abstract: Sovereign network functions, e.g., routing protocols, are becoming increasingly complex and susceptible to failures arising from protocol configuration anomalies and anomalous configurations. This paper interprets the protocol configuration anomaly detection problem as detection of structural inconsistencies of connected nodes and edges in a bipartite graph that captures both physical network entities and logical protocol states. This graph structural inconsistency detector (GSID) model is proposed to solve the problem efficiently. To handle the heterogeneous nature of protocol configuration parameters, GSID employs an adaptive configuration encoder (ACE) that dynamically selects encoding strategies per parameter to preserve fine-grained numerical discrepancies. To expose the subtle inconsistencies of connected nodes and edges in the bipartite graph, GSID uses an inconsistency dynamic attention (IDA) mechanism that scores edges by drawing asymmetric attentions from both ends, rule compliance from one end and route connectivity from the other. It is demonstrated experimentally that GSID outperforms state-of-the-art baselines by threefold in F1 score and by 23.2% in accuracy. Ablation studies validate the effectiveness of both the ACE and IDA modules. Tests on unseen network scales and real-world network topologies show the superior adaptability of our GSID, compared to the baselines.
arXiv:2603.03328v2 Announce Type: replace
Abstract: Language exhibits inherent structures, a property that explains both language acquisition and language change. Given this characteristic, we expect language models to manifest their own internal structures as well. While interpretability research has investigated how models compute representations mechanistically through attention patterns and Sparse AutoEncoders, the organization of the resulting representations is overlooked. To address this gap, we introduce StructLens, a framework to analyze representations through a holistic structural view. StructLens constructs maximum spanning trees based on the semantic representations in residual streams, inspired by tree representation in dependency parsing, and provides summaries of token relationships in representation space. We analyze how contiguous tokens are also nearby in representation space and find that middle layers show the strongest local-span organization. Moreover, analysis of pre-training checkpoints reveals that smaller local units become detectable earlier in pre-training, and larger units later. Our findings demonstrate that StructLens provides insights into how models organize token representations across layers and training. Our code is available at https://github.com/naist-nlp/structlens.
arXiv:2605.18475v1 Announce Type: new
Abstract: Mixed-precision quantization improves the budget--accuracy trade-off for large language models (LLMs) by allocating more bits to sensitive modules. However, automating this allocation at LLM scale faces a unique combination of constraints: learnable approaches require quantization-aware training, which is infeasible for billion-parameter models; training-free alternatives rely on static proxy metrics that miss cross-module interactions and must be recomputed per target budget; and search-based methods are expensive without guaranteeing exact budget compliance. We propose GAMMA, a quantizer-agnostic framework that learns module-wise precision preferences entirely within a post-training pipeline. GAMMA optimizes a teacher-forced hidden-state reconstruction objective under an augmented Lagrangian constraint, and projects the learned preferences into exact budget-feasible discrete assignments via integer programming. A key property is score reuse: because the learned preferences encode a stable sensitivity ranking rather than budget-specific weights, a single training run serves arbitrary deployment targets by re-solving only the integer program, reducing per-budget adaptation from hours to a few minutes. Across Llama and Qwen models (8B--32B), GAMMA outperforms both fixed-precision baselines (up to +12.99 Avg.) and search-based mixed-precision methods (up to +7.00 Avg.), and can match fixed 3-bit quality at 2.5-bit average precision, enabling deployment at substantially smaller memory footprints.
arXiv:2605.18527v1 Announce Type: new
Abstract: We present a systematic experimental comparison of single-pass second-harmonic generation (SHG) in bismuth triborate (BiBO) and lithium triborate (LBO) nonlinear crystals, driven by a 1.3 ps, 91 kHz laser at 1030 nm with up to 57 W of average input power. Both crystals yielded 32 W of second harmonic (SH) output at 515 nm, corresponding to a conversion efficiency of 56 %, which to the best of our knowledge represents the highest SH output power reported in the green spectral region using a BiBO crystal. Power dependence, long-term stability, beam quality, pulse duration, spectral properties, thermal effects, and angular acceptance bandwidth are characterized and directly compared for both crystals. These results provide quantitative performance benchmarks to guide the selection of nonlinear crystals for high-average-power, ultrashort-pulse frequency conversion near 1030 nm.
arXiv:2503.01696v3 Announce Type: replace
Abstract: We introduce and analyze a mesh-free two-level hybrid Chebyshev-Tucker tensor representation for approximating multivariate functions, which combines tensor-product Chebyshev interpolation with the low-rank Tucker decomposition of the tensor of Chebyshev coefficients. This construction allows to avoid the expensive rank-structured grid-based approximation of function-related tensors on large spatial grids, while benefiting from the Tucker decomposition of the moderate-sized core tensor of Chebyshev coefficients. Thus, we can compute the nearly optimal Tucker decomposition of the 3D function with controllable accuracy $\varepsilon >0$ without discretizing the function on a full fine grid in the domain, but only using its values at a small set of Chebyshev nodes computed either from the explicit analytic expression of the target function or from its data-sparse representation in a rank-structured tensor format with moderate rank parameter. Finally, we can represent the function in the algebraic Tucker format with optimal $\varepsilon$-rank on an arbitrarily large 3D tensor grid in the computational domain by discretizing the Chebyshev polynomials on that grid. The rank parameters of the nonlinear Tucker-ALS decomposition of the coefficient tensor can be much smaller than the polynomial degrees of the initial Chebyshev linear interpolation in the function independent polynomial basis set. It is shown that our techniques can be gainfully applied to the long-range part of the singular electrostatic potential of multi-particle systems represented on a fine grid in the range-separated (RS) tensor format. We provide error and complexity estimates and demonstrate the computational efficiency of the proposed techniques on challenging examples, including the collective electrostatic potential for large bio-molecular systems and lattice-type compounds.
arXiv:2605.17712v1 Announce Type: new
Abstract: Generative Artificial Intelligence (GenAI) has prompted significant discussion in education, yet large-scale empirical evidence on how students and teachers perceive and navigate this shift remains limited. We analyse 270k AI-related Reddit posts and comments from 26 education-related subreddits spanning higher education, K-12 teaching, and professional training between November 2022 and April 2026. Topic modelling reveals seventeen themes covering academic integrity, teaching & pedagogy, career anxiety, policy, and niche professional contexts. Discourse evolves from an early detection-and-evasion arms race into a sustained enforcement regime that constructive integration only begins to challenge in mid-2024. Stakeholder communities differ sharply: K-12 teachers foreground cognitive dependency, academics focus on AI detection and deliberation, and professional-programme students concentrate on career anxiety. Sentiment correlates strongly negatively with engagement, showing adversarial enforcement themes mobilise communities far more than constructive integration discourse. Examining where faculty and students meet, we find 17% of threads are cross-role, and one third of such contact occurs in the adversarial themes AI Detection and Misconduct Enforcement. Students initiate 68% of mixed threads, but faculty produce most cross-role replies. Mixed threads contain 2-3 times more records and last 2-4 times longer than same-role threads, making adversarial integrity disputes the center of sustained faculty-student contact. We discuss implications for governance, pedagogical design, and cross-role contact design. The code and data is available at https://github.com/tugrulz/genai-edu
arXiv:2605.03805v2 Announce Type: replace
Abstract: We develop component evolution (CE), a framework based on complex function theory for finite-blocklength channel polarization on discrete binary-input memoryless output-symmetric (BMS) channels. In this view, the Bhattacharyya parameter is treated as a real-valued instance of a broader class of complex-valued channel functionals. CE systematically derives analytic expressions for the Bhattacharyya parameters of the bit-channels of a given discrete BMS channel at arbitrary polarization levels. CE also enables structural analysis, providing new evidence of extremality of the binary erasure channel (BEC) and binary symmetric channel (BSC), and revealing new channel-dependent recursions for a class of BSC bit-channels.
arXiv:2605.18223v1 Announce Type: new
Abstract: Mineral dissolution in porous media is classically partitioned into static regimes within the Pe-Da plane, but this framework fails to capture the dissolution behavior of structurally complex rocks. Using three-dimensional micro-continuum simulations on micro-CT images of three rock samples spanning a wide range of pore-space heterogeneity, we track the joint evolution of dissolution morphology, velocity distribution, and reaction rate. Our results reveal that initial flow heterogeneity controls accessibility of reactants, thereby controlling the dissolution regime,reshaping them as dynamic trajectories. Channeled dissolution emerges as a simultaneous reorganization of structure and flow, and the resulting permeability-porosity relationship cannot be captured by a single power-law. The effective power-law exponent increases with heterogeneity and changes over time, reaching a maximum of 9.8, 18.0, and 40.9 for the three samples. Consequently, the effective reaction rate falls one to three orders of magnitude below the uniform dissolution prediction, with the suppression scaling with flow heterogeneity due to mass transfer limitations in channeled dissolution.
arXiv:2502.16691v2 Announce Type: replace
Abstract: Recent research has increasingly focused on training large language models (LLMs) using federated learning, known as FedLLM. However, responsible AI (RAI), which aims to ensure safe and trustworthy responses, remains underexplored in this context. In FedLLM, client-side training data may contain harmful content, resulting in unsafe LLMs that can generate inappropriate responses. Aggregating such models into a global model and redistributing it to clients risks the widespread deployment of unsafe LLMs. To address this, we incorporate two well-established RAI techniques into FedLLM: safety filtering and constitutional AI. Our experiments show that these methods significantly improve LLM safety, achieving over 20% improvement on AdvBench.
arXiv:2603.11276v2 Announce Type: replace-cross
Abstract: Real-world contextual bandit problems with complex reward models are often tackled with iteratively trained models, such as boosting trees. However, it is difficult to directly apply simple and effective exploration strategies--such as Thompson Sampling or UCB--on top of those black-box estimators. Existing approaches rely on sophisticated assumptions or intractable procedures that are hard to verify and implement in practice. In this work, we explore the use of an exploration-free (pure-greedy) action selection strategy, that exploits the randomness inherent in model fitting process as an intrinsic source of exploration. More specifically, we note that the stochasticity in cross-validation based regularization process can naturally induce Thompson Sampling-like exploration. We show that this regularization-induced exploration is theoretically equivalent to Thompson Sampling in the two-armed bandit case and empirically leads to reliable exploration in large-scale business environments compared to benchmark methods such as epsilon-greedy and other state-of-the-art approaches. Overall, our work reveals how regularized estimator training itself can induce effective exploration, offering both theoretical insight and practical guidance for contextual bandit design.
arXiv:2502.05462v2 Announce Type: replace
Abstract: We propose a real-time implementable motion planning framework for cooperative object transportation by nonholonomic mobile manipulator robots (MMRs) in dynamic environments. Our global planner finds a path from start to goal through the static, obstacle-free regions in the environment and generates a set of convex, static, obstacle-free regions around the path using a novel, fast, and computationally lightweight ellipse-based technique. We introduce a nonlinear Model Predictive Control (NMPC) based real-time implementable planning technique that jointly plans feasible motion for the mobile base and the manipulator's arm and generates a kinodynamic feasible, collision-free trajectory for cooperative object transportation. Simulation and hardware experiments validate the efficiency of our proposed planning framework.
arXiv:2407.17130v2 Announce Type: replace
Abstract: The mathematical formulation of sign-changing problems involves a linear second-order partial differential equation in the divergence form, where the coefficient can assume positive and negative values in different subdomains. These problems find their physical background in negative-index metamaterials, either as inclusions embedded into common materials as the matrix or vice versa. In this paper, we propose a numerical method based on the constraint energy minimizing generalized multiscale finite element method (CEM-GMsFEM) specifically designed for sign-changing problems. The construction of auxiliary spaces in the original CEM-GMsFEM is tailored to accommodate the sign-changing setting. The numerical results demonstrate the effectiveness of the proposed method in handling sophisticated coefficient profiles and the robustness of coefficient contrast ratios. Under several technical assumptions and by applying the \texttt{T}-coercivity theory, we establish the inf-sup stability and provide an a priori error estimate for the proposed method.
arXiv:2407.03175v2 Announce Type: replace
Abstract: This note demonstrates that we can stably recover all symmetric Toeplitz matrices $\pmb{X}_0\in\mathbb{R}^{n\times n}$ of rank at most $r$ from a number of rank-one subgaussian measurements on the order of $r\log^{2} n$ with an exponentially decreasing failure probability by employing a nuclear norm minimization program. Our approach utilizes descent cone analysis through Mendelson's small ball method with the Toeplitz constraint. The key ingredient is to determine the spectral norm of a random matrix with Toeplitz structure, which may be of independent interest. This improves upon earlier analyses and resolves the conjecture in Chen et al. (IEEE Transactions on Information Theory, 61(7):4034--4059, 2015).
arXiv:2204.01611v3 Announce Type: replace
Abstract: Inspired by the cognitive science theory, we explicitly model an agent with both semantic and episodic memory systems, and show that it is better than having just one of the two memory systems. In order to show this, we have designed and released our own challenging environment, "the Room", compatible with OpenAI Gym, where an agent has to properly learn how to encode, store, and retrieve memories to maximize its rewards. The Room environment allows for a hybrid intelligence setup where machines and humans can collaborate. We show that two agents collaborating with each other results in better performance than one agent acting alone.
arXiv:2509.07793v4 Announce Type: replace-cross
Abstract: How should well-being be prioritised in society, and what trade-offs are people willing to make between fairness and personal well-being? We investigate these questions using a stated preference experiment with a nationally representative UK sample (n = 300), in which participants evaluated life satisfaction outcomes for both themselves and others under conditions of uncertainty. Individual-level utility functions were estimated using an Expected Utility Maximisation (EUM) framework and tested for sensitivity to the overweighting of small probabilities, as characterised by Cumulative Prospect Theory (CPT). A majority of participants displayed concave (risk-averse) utility curves and showed stronger aversion to inequality in societal life satisfaction outcomes than to personal risk. These preferences were unrelated to political alignment, suggesting a shared normative stance on fairness in well-being that cuts across ideological boundaries. The results challenge use of average life satisfaction as a policy metric, and support the development of nonlinear utility-based alternatives that more accurately reflect collective human values. Implications for public policy, well-being measurement, and the design of value-aligned AI systems are discussed.
arXiv:2509.16255v2 Announce Type: replace-cross
Abstract: Purpose: To develop a deep learning method for the automatic segmentation of spinal nerve rootlets on various MRI scans. Material and Methods: This retrospective study included MRI scans from two open-access and one private dataset, consisting of 3D isotropic 3T TSE T2-weighted (T2w) and 7T MP2RAGE (T1-weighted [T1w] INV1 and INV2, and UNIT1) MRI scans. A deep learning model, RootletSeg, was developed to segment C2-T1 dorsal and ventral spinal rootlets. Training was performed on 76 scans and testing on 17 scans. The Dice score was used to compare the model performance with an existing open-source method. Spinal levels derived from RootletSeg segmentations were compared with vertebral levels defined by intervertebral discs using Bland-Altman analysis. Results: The RootletSeg model developed on 93 MRI scans from 50 healthy adults (mean age, 28.70 years $\pm$ 6.53 [SD]; 28 [56%] males, 22 [44%] females) achieved a mean $\pm$ SD Dice score of 0.67 $\pm$ 0.09 for T1w-INV2, 0.65 $\pm$ 0.11 for UNIT1, 0.64 $\pm$ 0.08 for T2w, and 0.62 $\pm$ 0.10 for T1w-INV1 contrasts. Spinal-vertebral level correspondence showed a progressively increasing rostrocaudal shift, with Bland-Altman bias ranging from 0.00 to 8.15 mm (median difference between level midpoints). Conclusion: RootletSeg accurately segmented C2-T1 spinal rootlets across MRI contrasts, enabling the determination of spinal levels directly from MRI scans. The method is open-source and can be used for a variety of downstream analyses, including lesion classification, neuromodulation therapy, and functional MRI group analysis.
arXiv:2503.13126v2 Announce Type: replace
Abstract: We study a variant of the Strang splitting for the time integration of the semilinear wave equation under the finite-energy condition on the torus $\mathbb{T}^3$. In the case of a cubic nonlinearity, we show almost second-order convergence in $L^2$ and almost first-order convergence in $H^1$. If the nonlinearity has a quartic form instead, we show an analogous convergence results, where the order is reduced by $1/2$ in both cases. To our knowledge these are the best convergence results available for the 3D cubic and quartic wave equations under the finite-energy condition. Our approach relies on continuous- and discrete-time Strichartz estimates. We also make use of the integration and summation by parts formulas to exploit cancellations in the error terms. Moreover, error bounds for a full discretization using the Fourier pseudo-spectral method in space are given. Finally, we discuss a numerical example indicating the sharpness of our theoretical results.
arXiv:2605.04375v2 Announce Type: replace
Abstract: To unleash the full potential of AI for Science, we must untether the agents from a purely digital environment. The agent's ability to control and explore in real-world labs is essential because the physical lab remains foundational to scientific discovery. While some tasks can be performed on a computer (e.g., data analysis, running simulated experiments), Eureka moments could occur at any time while operating lab instruments (e.g., when a scientist notices unexpected clues, intuition may prompt a real-time course change). Although autonomous labs are on the rise, which expose programmable APIs to control scientific instruments via software, bridging the gap between increasingly powerful AI agents and automated lab equipment requires innovation that draws insights from computer systems.
We propose a new paradigm called ``Experiment-as-Code (EaC) Labs,'' where a core concept is to encode experiments as declarative configurations that can be compiled down to device-level APIs. AI agents come up with hypotheses and experiments, written as an ensemble of declarative configurations. The systems layer performs program analysis, safety checks, resource assignment, and job orchestration. Finally, programmatic experimentation occurs via actuating the device APIs. This is a general stack that is science-, lab-, and instrument-independent, representing a novel synthesis across the physical, systems, and intelligence layers to unleash the next breakthrough in AI for Science.
arXiv:2605.16480v1 Announce Type: cross
Abstract: Molecules are graphs, but large language models~(LLMs) are usually asked to reason about them through linear strings. The most popular molecular representation, SMILES, compresses atoms, bonds, branches and rings into a compact sequence in which topology is implicit, forcing LLMs to reconstruct molecular structure before performing the requested chemical operation. Here we introduce MoleCode, an LLM-native, training-free, graph-explicit molecular language in which all molecular components are represented as typed entities with persistent identifiers and explicit relations. MoleCode makes molecular topology directly readable, editable and auditable within the language context, allowing an LLM to operate on structure rather than recover it from syntax. Across molecular reasoning, editing, generation and analysis tasks, this representational shift improves frontier LLMs most strongly when structural access is limiting: unfamiliar molecules, topology-sensitive operations, larger structures and repetitive polymers. It also changes how inference is allocated, replacing long reasoning traces devoted to implicit structural reconstruction with shorter, more chemically directed reasoning over explicit atoms and bonds. In molecular optimization, this enables localized, property-aligned edits that preserve structural similarity to the starting compounds. The same Subgraph--Node--Edge grammar extends beyond small molecules to polymers, Markush structures, mechanism-style transformations and interleaved scientific documents, including research articles and patent disclosures in which chemical information is distributed across text and images. These results suggest that the interface between scientific objects and LLMs should not treat structure as something to be decoded from text. When the object of reasoning is relational, the structure itself should be part of the language.
arXiv:2512.22473v5 Announce Type: replace-cross
Abstract: Transformers empirically perform precise probabilistic reasoning in carefully constructed ``Bayesian wind tunnels'' and in large-scale language models, yet the mechanisms by which gradient-based learning creates the required internal geometry remain opaque. We provide a complete first-order analysis of how cross-entropy training reshapes attention scores and value vectors in a transformer attention head. Our core result is an \emph{advantage-based routing law} for attention scores, \[ \frac{\partial L}{\partial s_{ij}} = \alpha_{ij}\bigl(b_{ij}-\mathbb{E}_{\alpha_i}[b]\bigr), \qquad b_{ij} := u_i^\top v_j, \] coupled with a \emph{responsibility-weighted update} for values, \[ \Delta v_j = -\eta\sum_i \alpha_{ij} u_i, \] where $u_i$ is the upstream gradient at position $i$ and $\alpha_{ij}$ are attention weights. These equations induce a positive feedback loop in which routing and content specialize together: queries route more strongly to values that are above-average for their error signal, and those values are pulled toward the queries that use them. We show that this coupled specialization behaves like a two-timescale EM procedure: attention weights implement an E-step (soft responsibilities), while values implement an M-step (responsibility-weighted prototype updates), with queries and keys adjusting the hypothesis frame. Through controlled simulations, including a sticky Markov-chain task where we compare a closed-form EM-style update to standard SGD, we demonstrate that the same gradient dynamics that minimize cross-entropy also sculpt the low-dimensional manifolds identified in our companion work as implementing Bayesian inference. This yields a unified picture in which optimization (gradient flow) gives rise to geometry (Bayesian manifolds), which in turn supports function (in-context probabilistic reasoning).
arXiv:2510.04006v2 Announce Type: replace
Abstract: Data-driven machine learning (ML) models are reshaping weather forecasting and have shown the potential to accelerate and surpass traditional physics-based approaches, leading to a second revolution in the field after data assimilation. However, most ML forecast models are trained with weighted variable-wise losses on rollout forecasts that neglect cross-variable and spatial error covariance induced by physical coupling, often yielding overly smooth and physically unrealistic long-range forecasts. To address this, we reformulate model training as a four-dimensional variational data assimilation (4DVar) problem that treats reanalysis data as imperfect observations. This enables the loss function to incorporate cross-variable error covariance structures that capture multivariate dependencies and their associated errors. In practice, we approximate this objective by computing the loss in an autoencoder-learned latent space of global atmospheric states. By encoding complex nonlinear couplings among atmospheric variables, this representation allows the high-dimensional, complex error covariance matrix in model space to be approximated as nearly diagonal in latent space, substantially simplifying implementation. We show that rollout training with latent-space constraints improves long-term forecast skill, while better preserving fine-scale structures and physical realism than the widely used model-space loss. Finally, we extend this framework to accommodate heterogeneous data sources, enabling the forecast model to be trained jointly on reanalysis and multi-source observations within a unified theoretical formulation.
arXiv:2512.05910v2 Announce Type: replace-cross
Abstract: The Brunovsky canonical form provides sparse structural representations that are beneficial for computational optimal control, yet existing methods fail to compute it reliably. We propose a technique that produces Brunovsky transformations with substantially lower construction errors and improved conditioning. A controllable linear system is first reduced to the staircase form via an orthogonal similarity transformation. We then derive a simple linear parametrization of the transformations yielding the unique Brunovsky form. Numerical stability is further enhanced by applying a deadbeat gain before computing system matrix powers and by optimizing the linear parameters to minimize condition numbers.
arXiv:2604.15950v2 Announce Type: replace
Abstract: Pancreatic ductal adenocarcinoma (PDAC) segmentation on contrast-enhanced CT is inherently ambiguous: inter-rater disagreement among experts reflects genuine uncertainty rather than annotation noise. Standard deep learning approaches assume a single ground truth, producing probabilistic outputs that can be poorly calibrated and difficult to interpret under such ambiguity. We present TwinTrack, a framework that addresses this gap through post-hoc calibration of ensemble segmentation probabilities to the empirical mean human response (MHR) -the fraction of expert annotators labeling a voxel as tumor. Calibrated probabilities are thus directly interpretable as the expected proportion of annotators assigning the tumor label, explicitly modeling inter-rater disagreement. The proposed post-hoc calibration procedure is simple and requires only a small multi-rater calibration set. It consistently improves calibration metrics over standard approaches when evaluated on the MICCAI 2025 CURVAS-PDACVI multi-rater benchmark.