---
title: Quant Letter No. 118: November 2025, Week 1
url: https://www.ml-quant.com/issues/2025-11-04/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
issue_date: 2025-11-04
---


# Quant Letter No. 118: November 2025, Week 1

Sent 2025-11-04. 101 items.

## arXiv

### Finance

- __[Partial-Information Markets & Divergent Beliefs](http://arxiv.org/abs/2511.01486v1)__: Shows how a mathematical model explains why prices and trader biases converge as traders get more information, and gives the best way to combine expert opinions to hunt for arbitrage. (2025-11-03, shares: 5) · https://www.ml-quant.com/papers/arxiv/2511.01486/
- __[Probabilistic Rule Layers for Financial Drift](http://arxiv.org/abs/2510.26627v1)__: Presents a simple, interpretable rule-based correction layer that detects and explains shifts in credit-risk scores, validated on 2008 mortgage data. (2025-10-30, shares: 4) · https://www.ml-quant.com/papers/arxiv/2510.26627/

### Miscellaneous

- __[Here are some shorter title options (no numbering): - Excess Growth - Excess Rate - Growth Excess - Surplus Growth - Overgrowth - Growth Surplus Recommended: Excess Growth (keeps meaning but is more concise).](http://arxiv.org/abs/2510.25740v1)__: - Excess Growth - Excess Rate - Growth Excess - Surplus Growth - Overgrowth - Growth Surplus Recommended: Excess Growth (keeps meaning but is more concise).: The paper proves that a central portfolio metric—the excess growth rate—can be exactly described using basic information‑theory ideas and a few natural axioms. In short, it shows that the extra growth a portfolio achieves is essentially an information quantity, so portfolio performance can be understood like information gain. (2025-10-29, shares: 4) · https://www.ml-quant.com/papers/arxiv/2510.25740/

### Historical Trending

- __[AI Data Economics](http://arxiv.org/abs/2510.24990v1)__: Lays out data economics: why data is special, documents AI training-data deals, proposes a hierarchy of data units, and lists key research questions. (2025-10-28, shares: 22) · https://www.ml-quant.com/papers/arxiv/2510.24990/
- __[Fed Quantiles](http://arxiv.org/abs/2510.24362v1)__: Models a central bank that maximizes a quantile (not expected) utility, linking hawkish/dovish leanings to that quantile and finding the Fed is mostly dovish. (2025-10-28, shares: 14) · https://www.ml-quant.com/papers/arxiv/2510.24362/
- __[Omniscient Lazy Investor](http://arxiv.org/abs/2510.24467v1)__: Shows a perfectly informed but slow trader has a single optimal waiting time between trades set by execution costs and the price path's roughness (Hurst/fractal), supported by theory and data. (2025-10-28, shares: 9) · https://www.ml-quant.com/papers/arxiv/2510.24467/
- __[National High‑Dosage Tutoring Spending](http://arxiv.org/abs/2510.24899v1)__: Machine learning on incomplete ESSER plans estimates U.S. school districts spent about $2.2 billion on high‑dosage tutoring during COVID. (2025-10-28, shares: 8) · https://www.ml-quant.com/papers/arxiv/2510.24899/
- __[Automation, Skills, and Inequality](http://arxiv.org/abs/2510.24923v1)__: Whether automation raises or lowers inequality depends on how workers’ skills relate to each other and to the technology, and the effect can reverse as the technology improves. (2025-10-28, shares: 6) · https://www.ml-quant.com/papers/arxiv/2510.24923/
- __[PEARL: Liquid Private‑Equity Replication](http://arxiv.org/abs/2510.23183v1)__: Liquid Private‑Equity Replication: Introduces PEARL, an AI method that reconstructs private‑equity returns from liquid assets using timing and leverage adjustments to better match quarterly PE benchmarks. (2025-10-27, shares: 4) · https://www.ml-quant.com/papers/arxiv/2510.23183/
- __[Regional and Worker Effects of Immigration](http://dx.doi.org/10.1086/739196)__: Following the same workers over time reveals separate effects of immigration on wages, employment, and occupational upgrading that repeated cross‑section studies miss. (2025-10-28, shares: 3) · https://www.ml-quant.com/papers/doi/10-1086-739196/

## RePEc

### Finance

- __[Volatile KSE-30 Stocks](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs43069-025-00421-4%3Bh%3Drepec%3Aspr%3Asnopef%3Av%3A6%3Ay%3A2025%3Ai%3A1%3Ad%3A10.1007_s43069-025-00421-4)__: Used ML and copula models to see how KSE‑30 stocks move with the market, forecast volatility, and build a lower‑volatility portfolio. (2025-11-04, shares: 10) · https://www.ml-quant.com/papers/repec/spr-snopef-v-6-y-2025-i-1-d-10-1007-s43069-025-00421-4/
- __[Adaptive Trading for Emerging Markets](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1186%2Fs40854-025-00754-3%3Bh%3Drepec%3Aspr%3Afininn%3Av%3A11%3Ay%3A2025%3Ai%3A1%3Ad%3A10.1186_s40854-025-00754-3)__: Built and back‑tested an ARMA‑GARCH automated trading system to protect emerging‑market portfolios during crises and reduce capital flight. (2025-11-04, shares: 9) · https://www.ml-quant.com/papers/repec/spr-fininn-v-11-y-2025-i-1-d-10-1186-s40854-025-00754-3/
- __[RiskParity with Fat Tails & Dynamic Correlations](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fdoi.org%2F10.1111%2Fjtsa.12792%3Bh%3Drepec%3Abla%3Ajtsera%3Av%3A46%3Ay%3A2025%3Ai%3A2%3Ap%3A353-377)__: Proposed a risk‑parity method using expected shortfall with regime‑switching, fat‑tailed returns to lower turnover and improve crisis performance. (2025-11-04, shares: 9) · https://www.ml-quant.com/papers/repec/bla-jtsera-v-46-y-2025-i-2-p-353-377/
- __[Sharpe Strategies & the Adaptive Market Hypothesis](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Farchive.conscientiabeam.com%2Findex.php%2F29%2Farticle%2Fview%2F4102%2F8464%3Bh%3Drepec%3Apkp%3Ateafle%3Av%3A12%3Ay%3A2025%3Ai%3A1%3Ap%3A120-142%3Aid%3A4102)__: Found Sharpe‑ratio minima/maxima trading strategies beat buy‑and‑hold across global indices (1998–2023), supporting adaptive investing. (2025-11-04, shares: 8) · https://www.ml-quant.com/papers/repec/pkp-teafle-v-12-y-2025-i-1-p-120-142-id-4102/
- __[Expectations Disagreement & Factor Premia](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0165176524006037%3Bh%3Drepec%3Aeee%3Aecolet%3Av%3A247%3Ay%3A2025%3Ai%3Ac%3As0165176524006037)__: Disagreement about future consumption explains market-wide returns, while disagreement about productivity mainly raises risk premia for small, low-profit firms. (2025-11-04, shares: 5) · https://www.ml-quant.com/papers/repec/eee-ecolet-v-247-y-2025-i-c-s0165176524006037/
- __[Pension Optimization under Jump-Diffusion](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0167668725000137%3Bh%3Drepec%3Aeee%3Ainsuma%3Av%3A121%3Ay%3A2025%3Ai%3Ac%3Ap%3A100-110)__: Derives closed-form rules for how a target‑benefit pension should invest and set benefits when returns have continuous and sudden (jump) risks. (2025-11-04, shares: 5) · https://www.ml-quant.com/papers/repec/eee-insuma-v-121-y-2025-i-c-p-100-110/
- __[Linked-Window DEA for Strategy Evaluation](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs10614-023-10528-7%3Bh%3Drepec%3Akap%3Acompec%3Av%3A65%3Ay%3A2025%3Ai%3A2%3Ad%3A10.1007_s10614-023-10528-7)__: Introduces a rolling DEA with linked decision units and Whale Optimization to find firms and trading strategies that stay efficient over time. (2025-11-04, shares: 5) · https://www.ml-quant.com/papers/repec/kap-compec-v-65-y-2025-i-2-d-10-1007-s10614-023-10528-7/
- __[Early-Exercise Premium in American Puts](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdx.doi.org%2F10.1287%2Fmnsc.2023.00440%3Bh%3Drepec%3Ainm%3Aormnsc%3Av%3A71%3Ay%3A2025%3Ai%3A2%3Ap%3A1824-1845)__: Allowing optimal early exercise changes the expected raw and delta‑hedged returns of American puts, altering some option‑anomaly results versus European puts. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/inm-ormnsc-v-71-y-2025-i-2-p-1824-1845/
- __[Sustainable-Return Measure for Long-Term Returns](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fhdl.handle.net%2F10.1080%2F0015198X.2024.2401765%3Bh%3Drepec%3Ataf%3Aufajxx%3Av%3A81%3Ay%3A2025%3Ai%3A1%3Ap%3A33-62)__: Defines a “sustainable return” (a withdrawal rate that preserves real capital) and shows that return order and reinvestment matter more than average returns for long-term outcomes. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/taf-ufajxx-v-81-y-2025-i-1-p-33-62/

### Statistical

- __[LLM News Sentiment & Stock Volatility](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0165176524006086%3Bh%3Drepec%3Aeee%3Aecolet%3Av%3A247%3Ay%3A2025%3Ai%3Ac%3As0165176524006086)__: For Dow firms (2019–2023), news sentiment—especially when labeled by GPT‑4—strongly predicts intraday stock volatility. (2025-11-04, shares: 6) · https://www.ml-quant.com/papers/repec/eee-ecolet-v-247-y-2025-i-c-s0165176524006086/
- __[Gamified Speech-Emotion Crowdsourcing](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.mdpi.com%2F2075-4698%2F15%2F3%2F54%2Fpdf%3Bh%3Drepec%3Agam%3Ajsoctx%3Av%3A15%3Ay%3A2025%3Ai%3A3%3Ap%3A54-%3Ad%3A1597256)__: JPlus is a gamified app that helps journalists collect emotionally labeled speech, boosting emotion-recognition models and user motivation. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/gam-jsoctx-v-15-y-2025-i-3-p-54-d-1597256/
- __[Competition & Earnings Volatility Driving Zero Leverage](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.mdpi.com%2F1911-8074%2F18%2F2%2F73%2Fpdf%3Bh%3Drepec%3Agam%3Ajjrfmx%3Av%3A18%3Ay%3A2025%3Ai%3A2%3Ap%3A73-%3Ad%3A1582023)__: Companies facing tougher product-market competition, especially with high earnings volatility, are more likely to adopt zero‑leverage (no-debt) policies. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/gam-jjrfmx-v-18-y-2025-i-2-p-73-d-1582023/

### Machine Learning

- __[Mixture AR models for multiscale Chinese finance](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fhdl.handle.net%2F10.1080%2F01605682.2024.2329156%3Bh%3Drepec%3Ataf%3Atjorxx%3Av%3A76%3Ay%3A2025%3Ai%3A1%3Ap%3A97-110)__: Proposes a new mixture-based autoregressive model that better captures complex financial time series and performs well during COVID-19. (2025-11-04, shares: 6) · https://www.ml-quant.com/papers/repec/taf-tjorxx-v-76-y-2025-i-1-p-97-110/
- __[Deep learning for VaR and ES estimation](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0305048324002135%3Bh%3Drepec%3Aeee%3Ajomega%3Av%3A133%3Ay%3A2025%3Ai%3Ac%3As0305048324002135)__: Develops two deep-learning frameworks (expectile and spline-quantile) plus a model-combination method to more accurately estimate Value-at-Risk and Expected Shortfall. (2025-11-04, shares: 6) · https://www.ml-quant.com/papers/repec/eee-jomega-v-133-y-2025-i-c-s0305048324002135/
- __[Quantile spillovers in international housing markets](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0261560625000014%3Bh%3Drepec%3Aeee%3Ajimfin%3Av%3A152%3Ay%3A2025%3Ai%3Ac%3As0261560625000014)__: Shows housing-price shocks transmit more strongly in the tails; the US is the main transmitter, and US interest rates strongly predict spillovers. (2025-11-04, shares: 5) · https://www.ml-quant.com/papers/repec/eee-jimfin-v-152-y-2025-i-c-s0261560625000014/
- __[Bond volatility as predictor of Chinese stock volatility](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fdoi.org%2F10.1002%2Ffor.3215%3Bh%3Drepec%3Awly%3Ajforec%3Av%3A44%3Ay%3A2025%3Ai%3A2%3Ap%3A547-555)__: Finds long-maturity Treasury futures volatility—especially the 10‑year—improves forecasts of Chinese stock volatility; SPCA and lasso beat HAR benchmarks and help investor returns. (2025-11-04, shares: 5) · https://www.ml-quant.com/papers/repec/wly-jforec-v-44-y-2025-i-2-p-547-555/
- __[Abstract journal classification: SVM vs BERT vs GPT-3.5](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs11192-024-05217-7%3Bh%3Drepec%3Aspr%3Ascient%3Av%3A130%3Ay%3A2025%3Ai%3A1%3Ad%3A10.1007_s11192-024-05217-7)__: SVM vs BERT vs GPT-3.5: Compares SVM, SPECTER, BERT, and GPT-3.5 for classifying abstracts: BERT performs best overall, while GPT-3.5 gives mixed results, especially with little data. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/spr-scient-v-130-y-2025-i-1-d-10-1007-s11192-024-05217-7/

### Deep Learning

- __[Polarized Spanish Anti‑Abortion Twitter Networks](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1057%2Fs41599-025-04568-7%3Bh%3Drepec%3Apal%3Apalcom%3Av%3A12%3Ay%3A2025%3Ai%3A1%3Ad%3A10.1057_s41599-025-04568-7)__: Study: Spanish-speaking anti-abortion Twitter accounts are mostly male, use hateful/negative messages, tied to religion/politics, and connect extreme-right voices. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/pal-palcom-v-12-y-2025-i-1-d-10-1057-s41599-025-04568-7/
- __[Demand Forecasting for New Fashion](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fdoi.org%2F10.1002%2Ffor.3192%3Bh%3Drepec%3Awly%3Ajforec%3Av%3A44%3Ay%3A2025%3Ai%3A2%3Ap%3A270-280)__: Review: Predicting demand for new fashion items is hard (changing tastes, seasonality, social media), but machine learning—especially deep learning and ensemble methods—can improve accuracy. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/wly-jforec-v-44-y-2025-i-2-p-270-280/
- __[Crisis Oil Price Forecasting: LSTM vs SVM](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs10479-023-05400-8%3Bh%3Drepec%3Aspr%3Aannopr%3Av%3A345%3Ay%3A2025%3Ai%3A2%3Ad%3A10.1007_s10479-023-05400-8)__: LSTM vs SVM: Comparison: During crises, LSTM (deep learning) predicts oil prices much better for mid-to-long-term horizons, while SVM performs poorly. (2025-11-04, shares: 4) · https://www.ml-quant.com/papers/repec/spr-annopr-v-345-y-2025-i-2-d-10-1007-s10479-023-05400-8/

### Historical Trending

- __[Adaptive VIX Forecasts](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fhdl.handle.net%2F10.1080%2F14697688.2024.2439458%3Bh%3Drepec%3Ataf%3Aquantf%3Av%3A24%3Ay%3A2024%3Ai%3A12%3Ap%3A1857-1873)__: Machine learning with dynamic nonlinear models and weekly jobless claims predicts daily VIX (market volatility) more accurately, helping forecasts and trading. (2024-02-09, shares: 12) · https://www.ml-quant.com/papers/repec/taf-quantf-v-24-y-2024-i-12-p-1857-1873/
- __[Eurozone Bank Stocks: DL vs Traditional](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.icfm.ro%2FRePEc%2Fvls%2Fvls_pdf%2Fvol28i4p29-42.pdf%3Bh%3Drepec%3Avls%3Afinstu%3Av%3A28%3Ay%3A2024%3Ai%3A4%3Ap%3A29-42)__: DL vs Traditional: For ten large Eurozone banks, traditional ML (XGBoost, logistic regression) outperformed LSTM/BiLSTM for daily stock direction—likely due to too little data for deep learning. (2024-01-06, shares: 7) · https://www.ml-quant.com/papers/repec/vls-finstu-v-28-y-2024-i-4-p-29-42/
- __[Gig Platform Work Satisfaction](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fjppc.ro%2Findex.php%2Fjppc%2Farticle%2Fdownload%2F868%2F470%3Bh%3Drepec%3Acta%3Ajcppxx%3A4241)__: The WNSS didn’t fit EU online gig workers; a simpler 12-item scale with three factors (survival, social contribution, competence) describes their needs better. (2024-08-05, shares: 2) · https://www.ml-quant.com/papers/repec/cta-jcppxx-4241/
- __[AI for Cloud Resources](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fnewjaigs.com%2Findex.php%2FJAIGS%2Farticle%2Fview%2F262%3Bh%3Drepec%3Adas%3Anjaigs%3Av%3A6%3Ay%3A2024%3Ai%3A1%3Ap%3A397-408%3Aid%3A262)__: AI methods (predictive analytics, reinforcement learning, anomaly detection) can automate and improve cloud resource management for more efficient DevOps. (2024-10-14, shares: 2) · https://www.ml-quant.com/papers/repec/das-njaigs-v-6-y-2024-i-1-p-397-408-id-262/
- __[Young Informal EU Workers (Pre-Pandemic)](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fdoi.org%2F10.2478%2Fjses-2024-0010%3Bh%3Drepec%3Avrs%3Ajsesro%3Av%3A13%3Ay%3A2024%3Ai%3A2%3Ap%3A82-97%3An%3A1005)__: Using 2019 Eurobarometer data, the paper profiles young informal workers in the EU27—their share, activities, sectors, and reasons—providing a pre-pandemic baseline. (2024-03-13, shares: 2) · https://www.ml-quant.com/papers/repec/vrs-jsesro-v-13-y-2024-i-2-p-82-97-n-1005/
- __[EGovernance & Participation](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fpublishing.globalcsrc.org%2Fojs%2Findex.php%2Fsbsee%2Farticle%2Fview%2F3089%2F1770%3Bh%3Drepec%3Asrc%3Asbseec%3Av%3A6%3Ay%3A2024%3Ai%3A3%3Ap%3A317-336)__: AIpowered e‑government can boost citizen participation but needs tech access, digital skills, trust, and good laws. (2024-04-27, shares: 2) · https://www.ml-quant.com/papers/repec/src-sbseec-v-6-y-2024-i-3-p-317-336/
- __[Dark Patterns in Investing](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.rivisteweb.it%2Fdownload%2Farticle%2F10.1435%2F115112%3Bh%3Drepec%3Amul%3Ajqmthn%3Adoi%3A10.1435%2F115112%3Ay%3A2024%3Ai%3A3%3Ap%3A499-531)__: Online dark patterns trick small investors; the paper calls for behavioral-science and AI-backed rules to stop them. (2024-04-12, shares: 1) · https://www.ml-quant.com/papers/repec/mul-jqmthn-doi-10-1435-115112-y-2024-i-3-p-499-531/
- __[Bank Performance Drivers](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fapi.eurokd.com%2FUploads%2FArticle%2F704%2FNCAF.2023.09.03.pdf%3Bh%3Drepec%3Abco%3Ancafaa%3A%3Av%3A9%3Ay%3A2023%3Ap%3A26-41)__: Bank performance depends on many factors, and COVID-19, digitalization, AI, and FinTechs are new major challenges. (2023-06-27, shares: 1) · https://www.ml-quant.com/papers/repec/bco-ncafaa-v-9-y-2023-p-26-41/
- __[AI Data Culture & Firm Performance](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs10796-023-10460-z%3Bh%3Drepec%3Aspr%3Ainfosf%3Av%3A26%3Ay%3A2024%3Ai%3A6%3Ad%3A10.1007_s10796-023-10460-z)__: Companies with strong AI infrastructure perform better, helped by a data-driven culture that also supports sustainability. (2024-09-07, shares: 1) · https://www.ml-quant.com/papers/repec/spr-infosf-v-26-y-2024-i-6-d-10-1007-s10796-023-10460-z/
- __[Social Media Climate Debate](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.cambridge.org%2Fcore%2Fproduct%2Fidentifier%2FS0027950124000073%2Ftype%2Fjournal_article%3Bh%3Drepec%3Acup%3Anierev%3Av%3A266%3Ay%3A2023%3Ai%3A%3Ap%3A22-29_3)__: Social-media ML shows clear, coordinated climate messaging across topics and policy levels is essential to reach net‑zero. (2023-04-23, shares: 1) · https://www.ml-quant.com/papers/repec/cup-nierev-v-266-y-2023-i-p-22-29-3/

## Machine learning

### Historical Trending

- __[Efficient 2D/3D Positional Encodings](https://arxiv.org/abs/2502.02562)__: STRING — a compact, translation-invariant position encoding that generalizes rotary encodings to any number of dimensions to boost vision and robotics models. (2025-02-04, shares: 23) · https://www.ml-quant.com/papers/arxiv/2502.02562/
- __[RiskAverse Conformal Prediction](https://arxiv.org/abs/2502.02561)__: RAC turns prediction sets into decision rules that provably maximize utility while keeping risk below a user-specified threshold. (2025-02-04, shares: 23) · https://www.ml-quant.com/papers/arxiv/2502.02561/
- __[COCONut-PanCap: Panoptic Segmentation & Grounded Captions](https://arxiv.org/abs/2502.02589)__: Panoptic Segmentation & Grounded Captions: COCONut-PanCap — a COCO-based dataset adding panoptic masks and dense region-level captions to improve vision–language understanding and generation. (2025-02-04, shares: 15) · https://www.ml-quant.com/papers/arxiv/2502.02589/
- __[AADDCE: Prostate DCE MRI GAN](https://arxiv.org/abs/2502.02555)__: Prostate DCE MRI GAN: AAD-DCE: an attention-based GAN that uses multiple MRI inputs to synthesize early and late DCE-MRI, reducing toxic contrast-agent use and outperforming prior methods. (2025-02-04, shares: 6) · https://www.ml-quant.com/papers/arxiv/2502.02555/
- __[Mosaic3D: Open-Vocabulary 3D Segmentation](https://arxiv.org/abs/2502.02548)__: Open-Vocabulary 3D Segmentation: Mosaic3D-5.6M: an automated pipeline and 3D mask–text dataset used to train Mosaic3D, achieving state-of-the-art open-vocabulary 3D semantic and instance segmentation. (2025-02-04, shares: 6) · https://www.ml-quant.com/papers/arxiv/2502.02548/
- __[EPV Benchmark & U-Net Pass Value](https://arxiv.org/abs/2502.02565)__: OJN-Pass-EPV: a new benchmark and U-Net EPV model (predicting ball height and pass risk/reward) that correctly identifies the higher-value game state about 78% of the time. (2025-02-04, shares: 5) · https://www.ml-quant.com/papers/arxiv/2502.02565/
- __[OMatG: Generative Model for Crystal Discovery](https://arxiv.org/abs/2502.02582)__: Generative Model for Crystal Discovery: OMatG: a generative framework using stochastic interpolants and symmetry-aware (equivariant) crystal representations to design stable inorganic crystals, setting a new state of the art. (2025-02-04, shares: 3) · https://www.ml-quant.com/papers/arxiv/2502.02582/

## Papers with code

### Trending

- __[Kimi Linear: Hybrid Linear Attention](https://github.com/MoonshotAI/Kimi-Linear)__: Hybrid Linear Attention: Kimi Linear: a new hybrid linear‑attention model that beats full attention while using far less memory, decoding much faster, and with code and checkpoints open‑sourced. (2025-10-31, shares: 500)
- __[FP Fixes Training–Inference Mismatch in RL Fine-tuning](https://github.com/sail-sg/Precision-RL)__: Switching RL fine‑tuning from BF16 to FP16: fixes numerical mismatches between training and inference, making training more stable and faster. (2025-11-03, shares: 88)

### Rising

- __[Toolathlon: Long‑Horizon Agent Benchmark](https://github.com/hkust-nlp/Toolathlon)__: Long‑Horizon Agent Benchmark: Toolathlon — a benchmark across 32 apps and 604 tools that tests agents on long, realistic workflows and shows current models often fail to complete them. (2025-10-30, shares: 33)
- __[CALM Continuous Next‑Vector Models](https://github.com/shaochenze/calm)__: CALM — a method that compresses multiple tokens into one continuous vector and predicts those vectors instead of tokens, cutting generation steps and making generation much faster. (2025-11-03, shares: 19)

## GitHub

### Finance

- __[Quant Resources](https://github.com/cybergeekgyan/Quant-Developers-Resources)__: Curated study materials and tools to help people prepare for quant and software-engineering jobs at trading firms, HFTs, and hedge funds. (2023-10-16, shares: 1333)
- __[Top Rust ML Libs](https://github.com/e-tornike/best-of-ml-rust)__: A ranked list of the best machine-learning libraries available in Rust, comparing their strengths. (2021-02-09, shares: 437)
- __[NOFX Trading OS](https://github.com/NoFxAiOS/nofx)__: NOFX proposes a next-generation, AI-powered trading operating system to automate and optimize trading workflows. (2025-10-28, shares: 6031)
- __[Paper Trading](https://github.com/etrobot/open-alpha-arena)__: An AI paper-trading project (inspired by nof1 Alpha Arena) that uses cctx to fetch market quotes for simulated trading. (2025-10-21, shares: 346)
- __[DeepFund](https://github.com/HKUSTDial/DeepFund)__: NeurIPS25 demo DeepFund Pilot showcasing deep-learning tools to help evaluate and manage fund investments. (2025-02-15, shares: 208)

### Trending

- __[TOONJSON](https://github.com/toon-format/toon)__: TOON — a compact JSON-like prompt format that roughly halves token use and includes spec, benchmarks, and reference code. (2025-10-22, shares: 8024)
- __[OA Interview](https://github.com/Leader-board/OA-and-Interviews)__: Guide — explains how online assessments and interviews work and how to prepare for them. (2021-11-24, shares: 2228)
- __[Agent CLI Mux](https://github.com/manaflow-ai/cmux)__: Cmux — runs multiple coding-agent CLIs (Claude, Codex, Gemini, etc.) in parallel so you can handle many tasks at once. (2025-07-25, shares: 398)
- __[PyStack](https://github.com/h2oai/pystacknet)__: No title provided — no content available to summarize. (2018-09-04, shares: 295)
- __[LG Drive](https://github.com/linagora/twake-drive)__: Opensource Google Drive alternative for storing and sharing files. (2023-03-23, shares: 1802)

## News

### Quantitative

- __[Winton launches US futures fund](https://www.hedgeweek.com/winton-expands-into-us-mutual-fund-market-with-managed-futures-launch/)__: Winton, a UK quantitative firm, entered the US mutual fund market by becoming adviser to the Altegris Futures Evolution Strategy Fund. (2025-10-30, shares: 5)
- __[Bain Capital buys stake in Innocap](https://www.hedgeweek.com/bain-capital-acquires-minority-stake-in-innocap-as-demand-for-hedge-fund-smas-surges/)__: Bain Capital bought a minority stake in Montreal’s Innocap to tap growing institutional demand for separately managed accounts. (2025-10-30, shares: 3)
- __[Hedge funds cut oil shorts after sanctions](https://www.hedgeweek.com/hedge-funds-slash-bearish-oil-positions-following-us-sanctions-on-russian-producers/)__: Hedge funds sharply reduced short bets on Brent crude after US sanctions on Russia’s top oil companies triggered a large, record unwind. (2025-11-03, shares: 3)
- __[Millennium eyes $5bn](https://www.hedgeweek.com/millennium-targets-5bn-for-new-private-markets-fund/)__: Millennium Management is raising $5 billion to start a new private‑markets fund, moving beyond its usual liquid strategies. (2025-10-31, shares: 2)
- __[Carlyle talent returns](https://www.efinancialcareers.com/news/private-equity-firm-carlyle-s-top-technologists-are-resurfacing-in-strange-places)__: People and companies are returning to banks, but they’re using them in new, different ways than before. (2025-10-30, shares: 2)
- __[UBS nears CS integration](https://www.efinancialcareers.com/news/ubs-has-only-scratched-the-surface-of-credit-suisse-s-legacy-technology)__: A long‑awaited announcement/launch has finally happened. (2025-10-29, shares: 1)

### Miscellaneous

- __[Ananym urges LKQ to divest Europe](https://www.hedgeweek.com/activist-ananym-renews-push-for-lkq-to-divest-european-operations/)__: Ananym Capital urges LKQ to sell its European unit to raise cash for big share buybacks. (2025-11-03, shares: 1)
- __[JPMorgan hires trader made MD at 27](https://www.efinancialcareers.com/news/max-lee-jain-global-jpmorgan)__: A hedge fund has closed after just a ten-month run. (2025-11-03, shares: 1)
- __[Hedge funds cover Magnificent Seven shorts](https://www.hedgeweek.com/hedge-funds-forced-to-cover-magnificent-seven-shorts/)__: Goldman says surprise strong earnings at five big tech firms forced hedge funds to buy back shares and cover shorts. (2025-11-03, shares: 1)
- __[OConnor Sale](https://www.hedgeweek.com/oconnor-sale-progressing-says-ubs-cfo/)__: UBS is proceeding with the sale of its O’Connor hedge fund unit to US brokerage Cantor Fitzgerald. (2025-10-30, shares: 1)
- __[Macquarie Picks](https://news.google.com/rss/articles/CBMiqAFBVV95cUxNc25paDdEVldmMG5mTUYyNXFxRHdnMjdRTFR4cXE2NWZoU3hYRi13MWZMdC1IZ1c5VlVWSzNvWGkzREljd3JBMlV0NVVWR1REUnFTbGJjU3d5LUpIOUh0TEJPU1dOLWxiRUgzaURvX1ZBMnpNRDBQbFBRTWJGQlUtS1gzWVRWMkZFS1JDSzhjZkR6MnBPb2tFaHRYWlhBNDhkY09ON2lIeHA?oc=5)__: Macquarie’s Melbourne Cup predictions narrowly picked the winner — it was right by a nose. (2025-11-04, shares: 0)
- __[Challenge](https://www.ft.com/content/8a6502c3-f244-4b61-880b-b20cf03299cc)__: No article was provided to summarize. (2025-11-02, shares: 0)
- __[Citi CTO: AI Cuts](https://www.efinancialcareers.com/news/citi-s-cto-says-these-are-the-kinds-of-jobs-he-s-eliminating-with-ai)__: AI Cuts: Citi published a compendium of AI use cases — a short guide to how AI can be applied in finance. (2025-10-29, shares: 0)

## Podcasts

### Quantitative

- __Managed Futures Diversifier__: Kathryn Kaminski: Managed futures had a big recent drawdown, but they still diversify portfolios and patient investors may be rewarded over time. (2025-10-31, shares: 7)
- __[Bullish Credit Puts](https://interactive-brokers-podcast.podbean.com/e/are-credit-put-spreads-the-smartest-bullish-play/)__: Market Chameleon guests: A credit put spread is shown as a way to limit downside risk, manage volatility, and earn steady income. (2025-10-30, shares: 5)
- __[LTCMA 30 Outlook](https://centerforinvestmentexcellence.podbean.com/e/ltcma-s-30th-anniversary-shifting-landscapes-and-silver-linings/)__: Tim Lintern (J.P. Morgan): The 30th Long‑Term Capital Market Assumptions give long‑run forecasts that help shape strategic asset allocation. (2025-10-31, shares: 5)
- __[VortexNet — AI](https://datascienceathome.podbean.com/e/why-ai-researchers-are-suddenly-obsessed-with-whirlpools-ep-293/)__: Researchers: VortexNet, a neural network inspired by whirlpool fluid dynamics, aims to fix vanishing gradients and better handle long‑range dependencies. (2025-10-30, shares: 4)
- __[Central Bank Moves](https://atanyrate.podbean.com/e/global-rates-trick-or-treating-with-central-banks/)__: Ipek Ozil & Khagendra Gupta: Recent and upcoming central bank actions are changing interest‑rate derivative pricing and market behavior. (2025-10-31, shares: 4)

### Related

- __[DB Buyouts](https://audioboom.com/posts/8797274)__: Lara Edmonstone‑West: DB pension schemes should prioritize buy‑ins or buyouts to guarantee members’ pensions instead of trying to extract surpluses. (2025-10-30, shares: 3)
- __[EM Fixed Income](https://atanyrate.podbean.com/e/em-fixed-income-not-spooked-yet/)__: JPMorgan strategists: A review of recent market moves and what they mean for risks and opportunities in emerging‑market fixed‑income. (2025-10-30, shares: 3)
- __[China LNG Peak (early 2030s)](https://atanyrate.podbean.com/e/global-commodities-china-s-lng-demand-to-peak-in-early-2030s/)__: Analysts on China LNG: China’s LNG demand will peak around 2032, and growing infrastructure could let it become a major global LNG trader. (2025-10-31, shares: 3)
- __[Short Seller's Mind](https://rss.com/podcasts/confessionsmm/2301216)__: Alex Temiz: Practical lessons on disciplined short selling, trading psychology, and the habits needed to survive as a trader. (2025-10-31, shares: 3)
- __[Macro & Credit Opportunities](https://chrt.fm/track/F81DEC/traffic.megaphone.fm/GLD5629810905.mp3)__: Tony Yoseloff: How to invest across different market cycles and where to find opportunity in alternative assets. (2025-10-31, shares: 3)

## Blogs

### Quantitative

- __[Asian Quant](https://quantnet.com/threads/understanding-the-quantitative-finance-industry-in-asia.46411/)__: Quant finance used to be concentrated in big global centers like New York and London. (2025-11-02, shares: 7)
- __[Risk vs. Portfolio](https://quantnet.com/threads/risk-versus-portfolio-management.9953/)__: Modern risk management grew out of work by a small group of New York quants between 1987 and 1993. (2025-10-29, shares: 4)
- __[NYC Mayor Markets](https://rajivsethi.substack.com/p/markets-and-the-mayors-race)__: A prediction market shows the NYC mayoral race as basically decided, and Bill Ackman objects to that signal. (2025-11-01, shares: 3)
- __[NYC Mayor Markets](https://rajivsethi.substack.com/p/markets-and-the-mayors-race)__: Critics warn Polymarket’s high odds for Zohran Mamdani can be pushed by small bets and might mislead. (2025-11-01, shares: 3)
- __[NYC Mayor Markets](https://rajivsethi.substack.com/p/markets-and-the-mayors-race)__: Observers question why anyone would buy Mamdani at 95% on Polymarket, suggesting the price could be manipulated. (2025-11-01, shares: 3)

### Related

- __[Connors R3](https://www.quantifiedstrategies.com/larry-connors-r3-strategy/)__: Tests Larry Connors' R3 strategy and reports how it performed on historical data. (2025-11-04, shares: 2)
- __[Neural-Net Indicator Pt.3](https://dekalogblog.blogspot.com/2025/11/expressing-indicator-in-neural-net-form.html)__: Uses a neural net to tune a trading indicator on 10‑minute EUR/USD and shows preliminary results. (2025-11-03, shares: 2)
- __[Neural-Net Indicator Pt.3](https://dekalogblog.blogspot.com/2025/11/expressing-indicator-in-neural-net-form.html)__: Follow‑up with additional neural‑net indicator tests and a discussion of what those results mean for trading. (2025-11-03, shares: 2)
- __[LongTerm Trading Strategies](https://www.quantifiedstrategies.com/3-long-term-strategies-that-work/)__: Outlines three long‑term trading strategies designed to reduce drawdowns and capture more bull‑market gains. (2025-11-03, shares: 2)
- __[Coding Traders](https://quantnet.com/threads/our-best-traders-spend-a-lot-of-their-time-pounding-away-writing-code.10630/)__: NYT piece on how technology and regulations like Dodd‑Frank have transformed trading on Wall Street. (2025-10-29, shares: 2)

## Reddit

### Quantitative

- __[Polygon API Test](https://www.reddit.com/r/algotrading/comments/1om8f9w/best_api_trying_polygonmassive_now/)__:  (2025-11-02, shares: 25)
- __[Summer '26 Quant Recruiting](https://www.reddit.com/r/quantfinance/comments/1onw7m0/summer_2026_intern_quant_trading_recruiting_cycle/)__:  (2025-11-04, shares: 26)

### Rising

- __[Month 1 Results](https://www.reddit.com/r/algotrading/comments/1okup18/my_first_month_live_results/)__:  (2025-10-31, shares: 109)
- __[Experienced Hires Switching Companies](https://www.reddit.com/r/quant/comments/1oksh7d/how_does_switching_companies_work_for_experienced/)__:  (2025-10-31, shares: 99)
- __[HRT AI](https://www.reddit.com/r/quant/comments/1oltuov/odd_lots_how_hudson_river_trading_actually_uses_ai/)__:  (2025-11-01, shares: 99)
- __[QT Comp — Amsterdam](https://www.reddit.com/r/quantfinance/comments/1olkqsw/qt_comp_in_amsterdam/)__:  (2025-11-01, shares: 32)
- __[Pitch Follow-Up](https://www.reddit.com/r/algotrading/comments/1onn75m/following_on_from_a_sales_pitch_earlier/)__:  (2025-11-03, shares: 41)