Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.23744

AlphaLedger: Portfolio-Grounded On-Policy Distillation for Regime-Aware Alpha Mining

Xiang Li·Zikai Wei·Yiyan Qi·Wanyun Zhou·Xiang Liu·Penglei Sun·Zhenheng Tang·Yongqi Zhang·Jian Guo·Xiaowen Chu
View original source ↗Read source document ↗Export metadata ↓

Abstract

Hosted LLMs enable automated alpha mining but incur recurring API costs and expose research context to external providers. Training a local generator requires supervision aligned with portfolio utility: a predictive factor may add little after redundancy and trading costs. This incremental value is assessed only after the factor program is complete, whereas generation proceeds sequentially. Deployment must also adapt to evolving stock-specific evidence and changes in factor effectiveness across market regimes. We propose AlphaLedger to address these challenges. Economic-context on-policy distillation (EC-OPD) conditions a frozen teacher on completed factors and their cost-adjusted marginal portfolio contributions, then distills its legal-action distributions along the student's own trajectories. The deployed generator requires neither teacher calls nor outcome inputs. EviPool constructs a dynamic stock universe from lagged multi-source evidence, while a regime-residual gate (RRG) reweights factors using matured, state-matched returns with shrinkage and bounded deviations from equal weighting. On Chinese A-shares, AlphaLedger achieves a 55.4\% annualized net return, a Sharpe ratio of 2.06, and a maximum drawdown of 10.8\%. In a common CSI500 universe, its annualized net return exceeds the strongest baseline by 43.6\% (31.6\% vs.\ 22.0\%), with a 13.5\% higher Sharpe ratio.

Explore related subjects

cs.CE

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Xiang Li, Zikai Wei, Yiyan Qi, Wanyun Zhou, Xiang Liu, Penglei Sun, Zhenheng Tang, Yongqi Zhang, Jian Guo, Xiaowen Chu. 2026-10-16. AlphaLedger: Portfolio-Grounded On-Policy Distillation for Regime-Aware Alpha Mining. https://arxiv.org/abs/2610.23744

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Revisiting Structure-Based Filtering for AI-Designed Antibodies: Opportunities, Limits, and What Matters

AI antibody design can generate far more candidates than can be experimentally tested, making post-generation filtering a central bottleneck. We study whether structure-prediction outputs can help prioritize AI-designed antibodies for experimental validation. We evaluate this problem on 4,184 experimentally characterized antibody-antigen pairs, combining newly generated wet-lab measurements with curated affinity data across diverse antigens and assay settings. We introduce two complementary filtering signals: Regional Interaction Contrast (RIC), which captures support for the predicted antibody-antigen interface, and Non-CDR Antibody Confidence (NAC), which measures antibody confidence outside the CDRs. Across AlphaFold 3, Protenix-v2, and ESMFold2, we find a striking difference in what information matters for filtering. Interface evidence is strongest when comparing candidates designed for different antigens, whereas antibody-level reliability is substantially more informative for ranking candidates within the same antigen. In particular, non-CDR antibody confidence consistently outperforms whole-antibody confidence across predictors, revealing an unexpected source of filtering signal beyond the predicted interface. Finally, we examine the information encoded in the predictor outputs themselves. Predicted structures contain substantial but incomplete binding-related signal, revealing both the opportunities and limits of structure-based filtering: better metrics can exploit current predictors more effectively, but their performance ultimately remains constrained by what those predictors can distinguish.

cs.CE↗
arXiv2026

Reinforcement Learning-Based Pricing and Operation for Hybrid Electric and Hydrogen Vehicle Stations

Electric vehicles (EVs) and hydrogen fuel cell vehi- cles (HFCVs) are promising technologies in the efforts to reduce carbon emissions. Challenges such as limited infrastructure, con- sumer dissatisfaction, and electrical grid strain inhibit progress in the adoption of these technologies. A zero emission vehicle (ZEV) station, capable of handling EVs and HFCVs, addresses these challenges by focusing on maximizing station profit and consumer utility, while minimizing grid impacts. The purpose of this research is to develop a realistic simulation of a ZEV station that is capable of handling real-time pricing decisions enabled by reinforcement learning (RL) and to compare the performance of the RL agent to static pricing methods. After training, the RL agent gains 10.83% greater station profit than the baseline manual parameter search. This work serves as a useful tool for ZEV station owners when handling complex pricing and purchasing decisions.

cs.CE↗
arXiv2026

BioXArena: Benchmarking LLM Agents on Multi-Modal Biomedical Machine Learning Tasks

Large language model (LLM) agents increasingly automate machine-learning model building, yet biomedical benchmarks emphasize question answering, reasoning, and tool use or cover narrow slices of biomedical ML coding. We introduce BioXArena, a biomedical machine learning (BioML) coding benchmark for task-specific model building on heterogeneous, often multi-modal datasets. It contains 76 end-to-end tasks across 9 domains: sequence, single-cell, structure, network biology, chemical biology, perturbation dynamics, phenotype--disease, imaging, and text-integrated tasks. Each task packages primary-source data with hidden labels, held-out graders, and biology-aware metrics on a unified 0-to-100 scale; agents write runnable code, train models, and predict private test samples. BioXArena emphasizes realistic data interfaces: most tasks require agents to integrate multiple input sources, including 46 tasks that combine distinct modality families. We compare 18 harness--backbone configurations: general-purpose coding, biomedical, and ML agents plus our Minimal runner, all under the same controlled resource configuration. Codex with GPT-6 Astra achieves the highest task-level average score of 73.91 among 18 configurations, closely followed by Claude Code with Opus 5.5 (73.79). Matched-backbone comparisons show higher observed scores for mature coding systems than for Minimal; seven Minimal backends characterize model differences under a fixed runner. Beyond the main leaderboard, ablations, robustness checks, and scaling, runtime, and failure-mode analyses characterize how backbones, scaffolds, budgets, and domains affect BioML coding performance, exposing modeling and execution gaps across agent and LLM backbone families.

cs.CE↗