---
title: Quant Letter No. 8: July 2023, Week 3
url: https://www.ml-quant.com/issues/2023-07-19/
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: 2023-07-19
---


# Quant Letter No. 8: July 2023, Week 3

Sent 2023-07-19. 97 items.

## arXiv

### Finance

- __[Order Book Dependent Hawkes Process for Large Datasets](https://arxiv.org/abs/2307.09077)__: A new high-frequency trading model uses a Hawkes process and high-dimensional functions from the order book, capable of handling billions of data points and tested on four NYSE stocks. (2023-07-18, shares: 8) · https://www.ml-quant.com/papers/arxiv/2307.09077/
- __[Machine Learning for Option Pricing Investigation](https://arxiv.org/abs/2307.07657)__: A study finds that the generalized highway network and a DGM variant improve the accuracy and training time of machine learning algorithms for option pricing. (2023-07-15, shares: 6) · https://www.ml-quant.com/papers/arxiv/2307.07657/
- __[Efficiency of Stock Price Prediction Models with Companies Projections](https://arxiv.org/abs/2307.07868)__: Researchers aim to create a model that uses company forecasts and sector performances to accurately predict short and long-term equity share prices. (2023-07-15, shares: 4) · https://www.ml-quant.com/papers/arxiv/2307.07868/
- __[Optimal Trade Execution Strategies under Fast Mean-Reversion](https://arxiv.org/abs/2307.07024)__: A study models market quality using uncertain volatility and liquidity, studying optimal strategy approximations and providing estimation methods for the model using high-frequency data. (2023-07-13, shares: 3) · https://www.ml-quant.com/papers/arxiv/2307.07024/
- __[PDEs for local stochastic vol. models](https://arxiv.org/abs/2307.09216)__: The article presents a new way to set prices in local stochastic volatility models, using rough path theory to understand conditional dynamics and price European options. (2023-07-18, shares: 3) · https://www.ml-quant.com/papers/arxiv/2307.09216/

### Economics

- __[Dynamic PCA for Forecasting with Many Predictors](https://arxiv.org/abs/2307.07689)__: A new dynamic forecasting method using supervised Principal Component Analysis (PCA) has been introduced, which is more effective in predicting U.S. macroeconomic variables. (2023-07-15, shares: 5) · https://www.ml-quant.com/papers/arxiv/2307.07689/
- __[Examining Trends in Interest Rates and Inflation](https://arxiv.org/abs/2307.08968)__: Research indicates that altering the start dates of economic samples by four years can significantly change profit growth trends, and high corporate profits don't necessarily lead to inflation. (2023-07-18, shares: 4) · https://www.ml-quant.com/papers/arxiv/2307.08968/
- __[Research Agenda: Datalism and Data Monopolies](https://arxiv.org/abs/2307.08049)__: Datalism and Data Monopolies: The emergence of data monopolies, firms that dominate data usage in their operations, is challenging the traditional concepts in Monopoly Capital Theory. (2023-07-16, shares: 3) · https://www.ml-quant.com/papers/arxiv/2307.08049/

### Crypto & Blockchain

- __[Real Economic Activities in the Bitcoin Blockchain Analysis](https://arxiv.org/abs/2307.08616?utm_source=dlvr.it&utm_medium=twitter)__: The research investigates the real economic activity in the Bitcoin blockchain involving retail users, raising concerns about Bitcoin ecosystem centralization and offering insights into the geographical location of Bitcoin users. (2023-07-17, shares: 4) · https://www.ml-quant.com/papers/arxiv/2307.08616/
- __[AMMs in Decentralized Prediction Markets](https://arxiv.org/abs/2307.08768)__: The article suggests a decentralized framework for prediction markets using automated market makers, with a focus on liquidity management and its impact on market behavior. (2023-07-17, shares: 5) · https://www.ml-quant.com/papers/arxiv/2307.08768/

### Historical Trending

- __[Fast Monte Carlo for Additive Processes and Option Pricing](https://arxiv.org/abs/2112.08291)__: The article introduces a quick Monte Carlo method for additive processes, improving accuracy in pricing options that depend on a specific path. (2021-12-15, shares: 27) · https://www.ml-quant.com/papers/arxiv/2112.08291/
- __[Averaging plus Learning Models and Asymptotics](https://arxiv.org/abs/1904.08131)__: The paper introduces unique models for agents interacting in financial markets and social networks, where unexpected events act as news, and agents learn from what they observe, offering fresh perspectives on social learning models. (2019-04-17, shares: 24) · https://www.ml-quant.com/papers/arxiv/1904.08131/
- __[Fast and Accurate Option Pricing with Deep Learning](https://arxiv.org/abs/2105.10467)__: New ultra-fast and highly accurate neural Transition Probability Density Function generators have been developed for use in computational finance. (2021-05-21, shares: 13) · https://www.ml-quant.com/papers/arxiv/2105.10467/

## SSRN

### Quantitative

- __[Sig-Splines: Time Series Generative Models Calibration](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4514421)__: Time Series Generative Models Calibration: A new model for analyzing multivariate time series data is proposed, using linear transformations and signature transforms instead of traditional neural networks, adding convexity to the model's parameters. (2023-07-18, shares: 46) · https://www.ml-quant.com/papers/ssrn/4514421/
- __[GANs and Synthetic Financial Data: VaR Calculation](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4512017)__: VaR Calculation: The article discusses the unique characteristics of financial data time series developed using a Generative Adversarial Neural net (GAN), emphasizing its applications in machine learning. (2023-07-16, shares: 3) · https://www.ml-quant.com/papers/ssrn/4512017/
- __[Risk Model of Machine Learning in Software Project Development](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4511875)__: The research identifies key risk factors causing failures in machine learning-based software projects, creating a risk model through literature review and expert surveys. (2023-07-15, shares: 2) · https://www.ml-quant.com/papers/ssrn/4511875/
- __[Machine Learning for Fake Job Detection](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4508792)__: The research introduces a machine learning technique to detect and halt fraudulent online job advertisements, safeguarding job hunters from scams. (2023-07-13, shares: 2) · https://www.ml-quant.com/papers/ssrn/4508792/
- __[Big Data Forecasting in SCM](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4515014)__: The article introduces a new framework for supply chain forecasting strategies and technologies, using Big Data Analytics for optimization and performance assessment. (2023-07-19, shares: 3) · https://www.ml-quant.com/papers/ssrn/4515014/
- __[Associations between Stock Price Volatility and Trading Frequency](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4514502)__: The study reveals that manipulated market volatility increases trading frequency, especially among non-high-risk gamblers, extending previous research on the link between disordered gambling and stock trading. (2023-07-17, shares: 2) · https://www.ml-quant.com/papers/ssrn/4514502/
- __[Value of Textual Data in Forecasting Macroeconomic Tail Risk](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4509043)__: News-based data offers valuable insights not provided by economic indicators, especially for left-tail forecasts, and significantly influences consumer sentiment. (2023-03-03, shares: 36) · https://www.ml-quant.com/papers/ssrn/4509043/
- __[Deep RL for Portfolio Optimization](https://arxiv.org/abs/2307.07694)__: The study finds that PPO and A2C deep reinforcement learning algorithms are more effective for portfolio optimization due to their noise handling and policy derivation capabilities, despite their high sample complexity. (2023-07-15, shares: 5) · https://www.ml-quant.com/papers/arxiv/2307.07694/
- __[Generative Meta-Learning for Portfolio Ensemble](https://arxiv.org/abs/2307.07811)__: The paper suggests a meta-learning method for creating a robust portfolio ensemble using a deep generative model, which balances sub-portfolio performance and correlation minimization, making it resilient to systematic shocks. (2023-07-15, shares: 5) · https://www.ml-quant.com/papers/arxiv/2307.07811/
- __[Portable Alpha for Taxable Investors](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4513238)__: Capital efficient retail products, like a 90/60 equity/bond strategy, can effectively replace long-only equity positions and help implement portable alpha strategies for taxable investors. (2023-02-15, shares: 2) · https://www.ml-quant.com/papers/ssrn/4513238/
- __[Equal Weight Index: Investing in Leading Companies Equally](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4512783)__: Investing in Leading Companies Equally: Standard & Poor's has launched the S&P Equal Weight Index, an investment opportunity that equally weighs the performance of 500 leading companies. (2023-02-17, shares: 3) · https://www.ml-quant.com/papers/ssrn/4512783/
- __[Reverse Causality: Credit Markets & Macroeconomic Shocks](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4511662)__: Credit Markets & Macroeconomic Shocks: Corporate bond credit spreads significantly react to macroeconomic shocks, with credit risk premia and leverage playing crucial roles, enhancing our understanding of credit markets and the macroeconomy. (2020-10-08, shares: 2) · https://www.ml-quant.com/papers/ssrn/4511662/
- __[ESG Investing: Factor-Tilt Approach](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4512638)__: Factor-Tilt Approach: A new portfolio construction method incorporates Environmental, Social, and Governance (ESG) factors, showing a significant positive ESG premium in the US market. (2023-05-25, shares: 2) · https://www.ml-quant.com/papers/ssrn/4512638/

### Financial

- __[Hedge Funds: With(out) Edge](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4513205)__: With(out) Edge: A new benchmark for assessing hedge fund performance is suggested, dividing funds into two groups based on their Sharpe ratios and skewness, and predicting their performance. (2023-07-17, shares: 3) · https://www.ml-quant.com/papers/ssrn/4513205/
- __[Sparse Modeling with Grouped Heterogeneity for Asset Pricing](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4511953)__: The paper presents a framework for clustering observations and selecting variables in panel data, aiming for economic interpretation and effective use of big data. (2023-07-15, shares: 8) · https://www.ml-quant.com/papers/ssrn/4511953/
- __[Fundamental Analysis in Stock Valuation Application](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4509690)__: The study finds that while fundamental analysis methods are useful in stock valuation, external factors like market sentiment and economic policy changes also influence stock prices. (2023-07-14, shares: 2) · https://www.ml-quant.com/papers/ssrn/4509690/
- __[Challenges in Sentiment Analysis for Investment Decisions](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4513208)__: Sentiment analysis can inform investment choices by interpreting sentiment from various text data sources, though it comes with its own challenges. (2023-07-17, shares: 2) · https://www.ml-quant.com/papers/ssrn/4513208/
- __[Bargaining and Choice in Intermediated Markets](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4512642)__: The article presents a theory examining the relationship between asset prices and liquidity in dealer-intermediated markets, suggesting it's nonmonotonic. (2023-07-17, shares: 5) · https://www.ml-quant.com/papers/ssrn/4512642/
- __[Dynamic Connectedness between Private Equities and High-Demand Assets](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4513308)__: Research shows a modest connection between a private equity ETF and high-demand asset classes, suggesting effective hedging through a short position in the ETF's volatility. (2023-07-17, shares: 2) · https://www.ml-quant.com/papers/ssrn/4513308/
- __[ETFs and Insider Trading](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4513419)__: Insider trading is hidden through 'shadow trading' in ETFs that include the target stock, with significant levels of such trading found before M&A announcements. (2023-02-01, shares: 2) · https://www.ml-quant.com/papers/ssrn/4513419/
- __[Machine Learning Alpha in Stock Return Predictions](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4514890)__: Machine learning models can predict stock returns effectively when trained on longer prediction periods and paired with efficient portfolio rules. (2023-06-18, shares: 2) · https://www.ml-quant.com/papers/ssrn/4514890/
- __[Liquidity Premium in Crypto Assets](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4514422)__: New ways of measuring liquidity premium Beta for crypto assets enhance predictability at high liquidity, outperforming traditional mean variance in portfolio performance. (2023-06-28, shares: 12) · https://www.ml-quant.com/papers/ssrn/4514422/
- __[Trustworthy Option Closing Prices: ML Approach](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4515200)__: ML Approach: A new machine learning model uses stock prices to predict options closing prices, outperforming traditional models and potentially improving market efficiency. (2023-03-01, shares: 2) · https://www.ml-quant.com/papers/ssrn/4515200/
- __[Linking Asset Prices to News without Mentions](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4508392)__: Semantic fingerprinting, a new Natural Language Processing technology, can better link news articles to asset prices by considering the overall text meaning. (2023-03-13, shares: 2) · https://www.ml-quant.com/papers/ssrn/4508392/

## RePEc

### Finance

- __[Portfolio Optimization with Factors](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0165176523001623%3Bh%3Drepec%3Aeee%3Aecolet%3Av%3A228%3Ay%3A2023%3Ai%3Ac%3As0165176523001623)__: A machine learning-based model that filters out noise from historical data enhances portfolio optimization by incorporating future-oriented information. (2023-07-19, shares: 23) · https://www.ml-quant.com/papers/repec/eee-ecolet-v-228-y-2023-i-c-s0165176523001623/
- __[Fama-French Model vs. Machine Learning](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.mdpi.com%2F2227-7390%2F11%2F13%2F2988%2Fpdf%3Bh%3Drepec%3Agam%3Ajmathe%3Av%3A11%3Ay%3A2023%3Ai%3A13%3Ap%3A2988-%3Ad%3A1186815)__: A seven-factor model, including the Hurst exponent and momentum factors, boosts the average R-squared by 7% in the A-share market, with SVM and random forests outperforming other machine learning algorithms. (2023-07-19, shares: 20) · https://www.ml-quant.com/papers/repec/gam-jmathe-v-11-y-2023-i-13-p-2988-d-1186815/
- __[VaR and ES Forecasting in Large Portfolios](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS2452306221000563%3Bh%3Drepec%3Aeee%3Aecosta%3Av%3A27%3Ay%3A2023%3Ai%3Ac%3Ap%3A1-15)__: Two new methods for estimating Value-at-Risk (VaR) and Expected Shortfall (ES) in large portfolios surpass existing methods, as per backtesting and scoring results. (2023-07-19, shares: 19) · https://www.ml-quant.com/papers/repec/eee-ecosta-v-27-y-2023-i-c-p-1-15/
- __[Factor Models for Large and Incomplete Data Sets: Unknown Structure](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0169207022000723%3Bh%3Drepec%3Aeee%3Aintfor%3Av%3A39%3Ay%3A2023%3Ai%3A3%3Ap%3A1205-1220)__: Unknown Structure: A new method that combines the expectation-maximization algorithm and the estimation algorithm for large factor models outperforms the standard EM algorithm in identifying clusters in large economic time series databases. (2023-07-19, shares: 14) · https://www.ml-quant.com/papers/repec/eee-intfor-v-39-y-2023-i-3-p-1205-1220/

### Machine Learning

- __[Explainable ML Methods for Actuarial Problems](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fwww.mdpi.com%2F2227-7390%2F11%2F14%2F3088%2Fpdf%3Bh%3Drepec%3Agam%3Ajmathe%3Av%3A11%3Ay%3A2023%3Ai%3A14%3Ap%3A3088-%3Ad%3A1193020)__: The article examines different explainable AI methods for data-based insurance issues, highlighting the need for accurate and understandable machine-learning solutions. (2023-07-19, shares: 24) · https://www.ml-quant.com/papers/repec/gam-jmathe-v-11-y-2023-i-14-p-3088-d-1193020/
- __[Internet Search Volume for Unemployment Predictions](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0169207022000656%3Bh%3Drepec%3Aeee%3Aintfor%3Av%3A39%3Ay%3A2023%3Ai%3A3%3Ap%3A1122-1144)__: The research suggests a mixed-frequency machine learning method using daily Google Trends data to predict weekly unemployment insurance claims, with improved accuracy during the COVID-19 crisis. (2023-07-19, shares: 20) · https://www.ml-quant.com/papers/repec/eee-intfor-v-39-y-2023-i-3-p-1122-1144/
- __[Choice Model vs. ML Techniques for Vehicle Ownership Decisions](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS0965856423001477%3Bh%3Drepec%3Aeee%3Atransa%3Av%3A173%3Ay%3A2023%3Ai%3Ac%3As0965856423001477)__: The piece explores the use of Machine Learning as an alternative to discrete choice models in planning, indicating that their predictive performance may differ based on context and comparison metrics. (2023-07-19, shares: 17) · https://www.ml-quant.com/papers/repec/eee-transa-v-173-y-2023-i-c-s0965856423001477/
- __[Improved Stock Market Volatility Prediction: New Bagging Model](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS1059056023001600%3Bh%3Drepec%3Aeee%3Areveco%3Av%3A87%3Ay%3A2023%3Ai%3Ac%3Ap%3A445-456)__: New Bagging Model: A model combining the autoregressive model and bagging method excels in predicting U.S. stock market volatility. (2023-07-19, shares: 14) · https://www.ml-quant.com/papers/repec/eee-reveco-v-87-y-2023-i-c-p-445-456/

### Deep Learning

- __[High-Frequency Trading Volume Prediction with Neural Networks](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Flink.springer.com%2F10.1007%2Fs11408-022-00421-y%3Bh%3Drepec%3Akap%3Afmktpm%3Av%3A37%3Ay%3A2023%3Ai%3A2%3Ad%3A10.1007_s11408-022-00421-y)__: A study successfully used a neural network to predict trading volumes of the CSI300 futures index using data from one to thirty minutes ahead, but found that adding data from nearby futures or spot trading volumes did not enhance the predictions. (2023-07-19, shares: 23) · https://www.ml-quant.com/papers/repec/kap-fmktpm-v-37-y-2023-i-2-d-10-1007-s11408-022-00421-y/
- __[Cement Tracking with Satellites and Neural Networks](https://econpapers.repec.org/scripts/redir.pf?u=https%3A%2F%2Fpublications.banque-france.fr%2Fsites%2Fdefault%2Ffiles%2Fmedias%2Fdocuments%2Fdt917.pdf%3Bh%3Drepec%3Abfr%3Abanfra%3A917)__: A new three-step machine learning method for predicting world trade has been proposed, which outperforms traditional linear, non-linear techniques and other benchmark models. (2023-07-19, shares: 18) · https://www.ml-quant.com/papers/repec/bfr-banfra-917/
- __[Hawkes Model Parameter Estimation with Recurrent Neural Networks](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fwww.sciencedirect.com%2Fscience%2Farticle%2Fpii%2FS1544612323002945%3Bh%3Drepec%3Aeee%3Afinlet%3Av%3A55%3Ay%3A2023%3Ai%3Apa%3As1544612323002945)__: A recurrent neural network has been used to estimate parameters of a Hawkes model using high-frequency financial data, showing faster computational performance and similar accuracy to traditional methods, allowing for real-time volatility measurement. (2023-07-19, shares: 14) · https://www.ml-quant.com/papers/repec/eee-finlet-v-55-y-2023-i-pa-s1544612323002945/

### Historical Trending

- __[Asset Volatility & Capital Structure in Corporate Mergers](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdx.doi.org%2F10.1287%2Fmnsc.2020.3607%3Bh%3Drepec%3Ainm%3Aormnsc%3Av%3A67%3Ay%3A2021%3Ai%3A5%3Ap%3A2773-2798)__: The study reveals that post-merger changes in leverage and cash holdings can be predicted by changes in asset volatility after corporate acquisitions. (2021-05-21, shares: 20) · https://www.ml-quant.com/papers/repec/inm-ormnsc-v-67-y-2021-i-5-p-2773-2798/
- __[Risk-Shifting & Volatility Puzzle](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdx.doi.org%2F10.1287%2Fmnsc.2020.3593%3Bh%3Drepec%3Ainm%3Aormnsc%3Av%3A67%3Ay%3A2021%3Ai%3A5%3Ap%3A2751-2772)__: The research indicates that shareholders face high unique risks when their companies are struggling, resulting in low or negative returns for firms with high unique volatility. (2021-02-16, shares: 15) · https://www.ml-quant.com/papers/repec/inm-ormnsc-v-67-y-2021-i-5-p-2751-2772/
- __[Long Memory & Fractality in Volatility](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdownloads.hindawi.com%2Fjournals%2Fcomplexity%2F2022%2F6728432.pdf%3Bh%3Drepec%3Ahin%3Acomplx%3A6728432)__: The study finds evidence of long memory and fractality in nine CBOE volatility indices, which could influence investment choices and trading tactics. (2022-03-25, shares: 14) · https://www.ml-quant.com/papers/repec/hin-complx-6728432/
- __[Adaptive Mixed-Attribute Clustering Method](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdownloads.hindawi.com%2Fjournals%2Fcomplexity%2F2022%2F6742120.pdf%3Bh%3Drepec%3Ahin%3Acomplx%3A6742120)__: A new data clustering method, AMDPC, based on density peaks, improves clustering accuracy by over 22.58% compared to the traditional K-prototype algorithm. (2022-01-08, shares: 14) · https://www.ml-quant.com/papers/repec/hin-complx-6742120/
- __[ML Approach for Predicting Pig Iron Production](https://econpapers.repec.org/scripts/redir.pf?u=http%3A%2F%2Fdx.doi.org%2F10.1287%2Finte.2020.1058%3Bh%3Drepec%3Ainm%3Aorinte%3Av%3A51%3Ay%3A2021%3Ai%3A3%3Ap%3A213-235)__: Machine learning is used to predict production levels in pig iron plants, aiding in improving efficiency and identifying key input variables. (2021-05-05, shares: 19) · https://www.ml-quant.com/papers/repec/inm-orinte-v-51-y-2021-i-3-p-213-235/

## Papers with code

### Trending

- __[GPTNeoX20B: OpenSource LM](https://github.com/labmlai/annotated_deep_learning_paper_implementations)__: OpenSource LM: The article introduces GPTNeoX20B, a language model with 20 billion parameters, trained on the Pile, and announces that its code will be made publicly available. (2023-07-15, shares: 29430)
- __[Llama 2: Foundation & Chat Models](https://github.com/facebookresearch/llama)__: Foundation & Chat Models: The piece discusses the creation and launch of Llama 2, a series of large language models with parameters ranging from 7 billion to 70 billion. (2023-07-19, shares: 27950)
- __[RetrievalAugmented Generation for NLP](https://github.com/deepset-ai/haystack)__: The article explores how large pretrained language models store factual information and achieve excellent results when fine-tuned for specific natural language processing tasks. (2023-07-16, shares: 9987)
- __[Petals: Collaborative Inference of Large Models](https://github.com/bigscience-workshop/petals)__: Collaborative Inference of Large Models: The article discusses the limitations of offloading and APIs in terms of speed and flexibility, particularly for research that requires access to weights, attention, or logits. (2023-07-19, shares: 5569)
- __[FlashAttention2: Faster Attention, Better Parallelism](https://github.com/dao-ailab/flash-attention)__: Faster Attention, Better Parallelism: The piece notes that inefficient GPU work partitioning can lead to low occupancy or unnecessary shared memory reads/writes. (2023-07-19, shares: 5014)

### Rising

- __[Secrets of RLHF in Large Language Models: PPO](https://github.com/openlmlab/moss-rlhf)__: PPO: The article discusses the use of PPOmax, an upgraded version of the PPO algorithm, to improve the stability of policy model training. (2023-07-13, shares: 323)
- __[InternLM: Evaluating Multimodal Models](https://github.com/InternLM/opencompass)__: Evaluating Multimodal Models: The authors introduce MMBench, a new benchmark designed to tackle challenges in multimodality. (2023-07-14, shares: 235)

## GitHub

### Finance

- __[Game Theory for ML Models](https://github.com/shap/shap)__: The article investigates how game theory can be used to understand the outcomes of machine learning models. (2016-11-22, shares: 19720)
- __[Timeseries ML at Scale](https://github.com/descendant-ai/functime)__: The piece explores how machine learning can be applied to analyze large sets of time series data and embeddings. (2023-06-05, shares: 280)
- __[RL for Market Making Strategies](https://github.com/KodAgge/Reinforcement-Learning-for-Market-Making)__: The article studies the use of tabular and deep reinforcement learning methods to identify optimal market strategies. (2022-05-20, shares: 59)
- __[TradingView Webhook Bot](https://github.com/lth-elm/TradingView-Webhook-Trading-Bot)__: The article outlines a Flask app that automates trading orders and sends trading charts to Discord via a bot. (2021-08-16, shares: 182)

### Trending

- __[GPT Researcher: Online Research Agent](https://github.com/assafelovic/gpt-researcher)__: Online Research Agent: An AI agent uses GPT technology to perform detailed online research on any given topic. (2023-05-12, shares: 918)
- __[ShortGPT: AI Framework for Short Video Creation](https://github.com/RayVentura/ShortGPT)__: AI Framework for Short Video Creation: ShortGPT is an AI system that automates the production of short video content, assisting creators in various aspects. (2023-06-27, shares: 749)
- __[Danswer: Private Source Backed Q&A](https://github.com/danswer-ai/danswer)__: Private Source Backed Q&A: A tool that enables users to ask questions in everyday language and get answers from private sources, compatible with platforms like Slack, GitHub, and Confluence. (2023-04-27, shares: 1712)
- __[Sweep: AI Junior Developer](https://github.com/sweepai/sweep)__: AI Junior Developer: Sweep is an AI system developed to operate as a junior software developer. (2023-06-14, shares: 1013)

## News

### Quantitative

- __[AI Benefits in Investment](https://news.google.com/rss/articles/CBMiemh0dHBzOi8vd3d3Lmdsb2JhbGludmVzdG9yZ3JvdXAuY29tL2FydGljbGVzLzM3MDA1NjUvYXNzZXNzaW5nLXRoZS1iZW5lZml0cy1vZi1haS1iYXNlZC1xdWFudGl0YXRpdmUtaW52ZXN0bWVudC1zdHJhdGVnaWVz0gEA?oc=5)__: The article assesses the benefits of integrating artificial intelligence into quantitative investment approaches. (2023-07-18, shares: 4)
- __[Quant Two Sigma Combines AI and Human Touch](https://news.google.com/rss/articles/CBMiVWh0dHA6Ly93d3cuaGVkZ2V3ZWVrLmNvbS8yMDIzLzA3LzE3LzMyMTMzNC9xdWFudC10d28tc2lnbWEtZW1wbG95LWh1bWFuLXRvdWNoLXRyYWRpbmfSAQA?oc=5)__: The piece explores Quant Two Sigma's choice to include human involvement in their trading operations. (2023-07-17, shares: 4)
- __[Data Science Jobs in Finance: Talent Wasted](https://www.efinancialcareers.com/news/2023/07/data-science-jobs-in-finance)__: Talent Wasted: The article investigates the difficulties in finding proficient data scientists in today's job market. (2023-07-18, shares: 3)
- __[AI Finance](https://news.google.com/rss/articles/CBMibmh0dHBzOi8vd3d3LmJsb29tYmVyZy5jb20vbmV3cy9hcnRpY2xlcy8yMDIzLTA3LTE3L2FpLWNhbi13cml0ZS1idXQtaXMtaXQtYW55LWdvb2QtYXQtcGlja2luZy1zdG9ja3MtcXVpY2t0YWtl0gEA?oc=5)__: AI technologies like ChatGPT and Google’s Bard are now being used in the finance sector, as reported by Bloomberg. (2023-07-17, shares: 2)
- __[AI's Stock Picking Skills Analyzed](https://news.google.com/rss/articles/CBMingFodHRwczovL3d3dy53YXNoaW5ndG9ucG9zdC5jb20vYnVzaW5lc3MvMjAyMy8wNy8xNy9haS1jYW4td3JpdGUtYnV0LWlzLWl0LWFueS1nb29kLWF0LXBpY2tpbmctc3RvY2tzLXF1aWNrdGFrZS85ZjViY2I1Ni0yNDZmLTExZWUtOTIwMS04MjZlNWJiNzhmYTFfc3RvcnkuaHRtbNIBAA?oc=5)__: The Washington Post examines the efficiency of AI in choosing stocks. (2023-07-17, shares: 0)

## Podcasts

- __[Potential Comeback of Bonds](https://chrt.fm/track/E5A66E/pdst.fm/e/rss.art19.com/episodes/a15b16a8-5cb8-4b5d-98fa-fc643825d886.mp3?rss_browser=BAhJIgtTYWZhcmkGOgZFVA%3D%3D--e8daa48e4e049c2293a0ad1663b4a762c475e386)__: The decrease in inflation is making government bonds more appealing, potentially leading to a resurgence later this year. (2023-07-14, shares: 8)
- __[Barry Ritholtz Interviews Thomas Wagner on Sports Investments](https://omny.fm/shows/masters-in-business/thomas-wagner-on-sports-investing-with-tom-brady)__: Thomas Wagner, co-founder of Knighthead Capital Management, talks about his sports investments and his finance career. (2023-07-14, shares: 6)
- __[Greenig: Trend Following in Alternative Markets](https://www.flirtingwithmodels.com/2023/07/18/s6e11-at-the-frontier-of-trend-following/)__: Trend Following in Alternative Markets: Doug Greenig talks about the unique features of alternative markets and their diversification potential in a podcast. (2023-07-17, shares: 5)

## Videos

### Quantitative

- __[Creating Private PDF ChatBot](https://www.youtube.com/watch?v=hSQY4N1u3v0)__: The tutorial video teaches how to create a private PDF chatbot using falcon7b and falcon40b models, without the need for abstract libraries. (2023-07-16, shares: 2)
- __[Deploying LLMs with One Click](https://www.youtube.com/watch?v=hSQY4N1u3v0)__: The tutorial video instructs on personalizing a PDF chatbot using falcon7b and falcon40b models, eliminating the need for abstract libraries. (2023-07-15, shares: 2)
- __[Deploying Private & Fast LLM Chatbots](https://www.youtube.com/watch?v=o1BCq1KJULM)__: The tutorial video guides on how to deploy and operate large language model chatbots locally using textgenerationinference and chatui, suitable for a production environment. (2023-07-14, shares: 48)

## X / Twitter

### Quantitative

- __[Poor Predictors: Analyst Price Targets](https://twitter.com/quantseeker/status/1679421427349372934)__: Analyst Price Targets: The article suggests that sell-side analysts' price targets are generally not reliable predictors of stock returns, but within-analyst demeaned/ranked price targets can predict unexplained returns. (2023-07-13, shares: 2)
- __[Operational Edge: Uncorrelated Markets](https://twitter.com/quantseeker/status/1681659769973612546)__: Uncorrelated Markets: The article emphasizes the need for adding new uncorrelated markets and maintaining simple portfolio construction and trend signals, a strategy known as operational edge. (2023-07-19, shares: 1)
- __[Synthetic Data & AI Models](https://twitter.com/quantseeker/status/1681564270235127814)__: The article explores the application of synthetic data in AI models. (2023-07-19, shares: 1)
- __[Forensic Finance Review](https://twitter.com/quantseeker/status/1680621681805598723)__: The article reviews Forensic Finance, covering areas like market manipulation, corporate fraud, insider trading, corruption, and greenwashing. (2023-07-16, shares: 2)

### Miscellaneous

- __[Python Time Series Forecasting Library with Papers](https://twitter.com/carlcarrie/status/1680549684073357312)__: A new Python library has been created for time series forecasting, incorporating the latest methods and citing relevant academic papers. (2023-07-16, shares: 1)
- __[Fast and Accurate Text Classification with K-means](https://twitter.com/carlcarrie/status/1679822034493112321)__: A novel text classification model, using a k-means classifier and gzip compression algorithm, is faster yet as precise as deep neural networks. (2023-07-14, shares: 1)
- __[Nevergrad: Gradient-free Optimization Toolbox in Python](https://twitter.com/carlcarrie/status/1681446463191285763)__: Gradient-free Optimization Toolbox in Python: Nevergrad, a Python toolbox, is specifically designed to carry out gradient-free optimization. (2023-07-19, shares: 0)
- __[Minimally Invasive Abstractions for LLMs Error Handling](https://twitter.com/carlcarrie/status/1679986304115781632)__: There is a demand for less intrusive abstractions around LLVM's error handling and tools, excluding other Python libraries and current idiomatic conventions. (2023-07-14, shares: 0)

## Reddit

### Quantitative

- __[YouTube Channels for Quant Research](https://www.reddit.com/r/quant/comments/14y58ik/recommend_yt_channels/)__:  (2023-07-13, shares: 19)
- __[Career Tips](https://www.reddit.com/r/quant/comments/14yi418/career_advice/)__:  (2023-07-13, shares: 19)
- __[Transitioning from Insurance Trading Desk](https://www.reddit.com/r/quant/comments/14zwxyi/advice_on_switching_from_insurance_trading_desk/)__:  (2023-07-15, shares: 9)
- __[Python Backtesting with Returns](https://www.reddit.com/r/quant/comments/151127e/how_do_you_guys_implement_returns_in_backtests_py/)__:  (2023-07-16, shares: 9)
- __[Latest on finqual Python Package](https://www.reddit.com/r/quant/comments/151m3k8/finqual_python_package_to_simplify_fundamental/)__:  (2023-07-17, shares: 29)

### Rising

- __[Quant Finance School & Career Paths](https://www.reddit.com/r/quant/comments/151autn/where_did_you_go_to_school_and_where_are_you_now/)__:  (2023-07-16, shares: 52)
- __[Understanding Citadel LLC's DSG Data Strategies](https://www.reddit.com/r/quant/comments/150i5d8/what_is_the_dsg_data_strategies_group_at_citadel/)__:  (2023-07-15, shares: 23)
- __[Role of a Power Trader Quant Explored](https://www.reddit.com/r/quant/comments/150s96v/what_dkes_a_power_trader_quant_do_actually/)__:  (2023-07-16, shares: 30)
- __[Real-World Exotic Derivatives Examples](https://www.reddit.com/r/quant/comments/14z32ov/what_are_exotic_derivatives/)__:  (2023-07-14, shares: 24)
- __[Breaking into Quant Roles at Lesser-Known Firms](https://www.reddit.com/r/quant/comments/14yt40i/finding_positions_at_lesserknown_companies/)__:  (2023-07-13, shares: 21)