Quant Letter: November 2023, Week 2

Weekly quantitative finance newsletter

ArXiv

Finance

Maximizing Portfolio Predictability with Machine Learning: Portfolio Predictability Maximization using ML: A stock portfolio called the maximally predictable portfolio (MPP), created using machine learning and a Kelly criterion strategy, consistently performs better than the benchmark. (2023-11-03, shares: 5)

Arbitrage Opportunities in Mean Field System: The article presents a theoretical model to analyze arbitrage opportunities in a market with unlimited investors, confirming the existence of a unique mean field equilibrium. (2023-11-05, shares: 3)

Transfer Risk and Finance Applications: The paper discusses the concept of transfer risk in transfer learning, showing its significant relation with performance and its effectiveness in selecting suitable source tasks in stock return prediction and portfolio optimization. (2023-11-06, shares: 2)

Power Law in Sandwiched Volterra Volatility Model: Power Law in Volterra Volatility Model: The Sandwiched Volterra Volatility (SVV) model accurately reproduces the power-law behavior of the at-the-money implied volatility skew, provided the correct Volterra kernel is chosen. (2023-11-02, shares: 4)

Optimal Stopping Problem with Discontinuous Reward: The study investigates the optimal stopping issue in pricing a variable annuity contract, introducing new valuation algorithms and showing how fee and surrender charge functions affect early and optimal surrender boundaries. (2023-11-06, shares: 2)

Joint Model for Longitudinal and Spatio-Temporal Survival Data: Longitudinal and Spatio-Temporal Survival Model: The Spatio-Temporal Joint Model (STJM) is a new method for credit risk analysis that uses spatial and temporal data to predict a borrower's risk, showing better results when spatial data is included. (2023-11-07, shares: 7)

Miscellaneous

Finding Fraud Prevention Rules: The paper introduces PORS, a heuristic-based framework for finding high-quality rule subsets in fraud prevention, and SpectralRules, a new sequential covering algorithm, showcasing their effectiveness in two real Alipay scenarios. (2023-11-02, shares: 4)

Asset Price Bubbles: Nonstationary Phenomenon: The article discusses the theory of rational asset price bubbles, highlighting that bubbles linked to real assets like stocks and housing are nonstationary phenomena tied to unbalanced growth. (2023-11-07, shares: 4)

Decentralization in Blockchain Governance and DeFi Efficiency: The article studies how decentralization in blockchain-based governance affects the financial efficiency of Decentralized Autonomous Organizations (DAOs). It uses the Gini coefficient to measure inequality among token owners and discusses the pros and cons of this method. (2023-11-04, shares: 4)

Historical Trending

Deep Learning for Volatility Calibration: The paper presents a new algorithm that uses deep self-consistent learning for better and more robust calibration of local volatility from market option prices. (2021-12-09, shares: 15)

Wage-Setting and Behavioral Firms: The study suggests that companies that set salaries at round numbers, typically less sophisticated firms, tend to perform worse in the market due to their coarse wage-setting approach. (2022-06-02, shares: 125)

Pragmatic Energy Markets: The article offers a guide on using the Heath-Jarrow-Morton framework in energy markets, specifically in European power and gas markets, covering market structure, model calibration, simulations, and derivatives pricing. (2023-05-02, shares: 56)

Multimodal Bankruptcy Prediction: The research presents multimodal learning in bankruptcy prediction models to tackle the problem of missing MDA section in Form 10-K, showing improved classification performance and addressing the limitation of previous models. (2022-10-26, shares: 33)

Liquidation with High Risk Aversion: The research investigates the Bachelier model with linear price impact, identifying a set of portfolios that are optimally effective in a scenario of diminishing price impact. (2023-01-04, shares: 10)

SSRN

Recently Published

Financial

Concave Price Impact Trading: The research examines statistical arbitrage issues, taking into account the nonlinear and temporary price impact of metaorders, and shows that simple trading rules can be established even with nonparametric alpha and liquidity signals. (2023-11-06, shares: 120.0)

Volatility Disagreement Trading: A model is created to understand how investors' disagreement on future volatility affects their trading of volatility derivatives, showing that trading decreases in more volatile periods and the variance risk premium can become positive when future volatility is underestimated. (2023-11-06, shares: 3.0)

Global Macro and Managed Futures Hedge Fund Strategies: The research evaluates the performance of hedge funds, especially those using a top-down investment approach, and discovers a significant drop in risk-adjusted alpha for global macro managers and managed futures managers after the global financial crisis. (2023-11-07, shares: 8.0)

Market Volatility and Trend Factor: The paper explores the link between stock market volatility and trend factor profits, finding that the trend factor performs better after high volatility periods as investors depend more on trend signals. (2023-11-02, shares: 3.0)

The Halo Effect in ESG Investing in Indian Equities: A study of 700 Indian companies shows no significant link between ESG scores and investment returns from 2013 to 2023. (2023-11-06, shares: 10.0)

The Kelly Criterion in Stock Investment: A paper suggests using the Kelly criterion and Monte Carlo simulation to estimate the optimal portfolio in stock investment. (2023-11-07, shares: 7.0)

Strategic Investors and Exchange Rate Dynamics: A study shows that exchange rate dynamics are affected by the diversity of investors and their price impact, with more concentrated markets having a stronger price impact. (2023-11-02, shares: 3.0)

Quantitative

Leverage Effect and Volatility of Volatility Estimation: The article presents new methods for estimating leverage effect and volatility using high frequency data, tested through simulation and real data analysis. (2023-11-07, shares: 3.0)

Machine Learning for Insolvency Prediction in Insurance: A new machine learning algorithm, SANN, is used to predict insurance company insolvency, showing better accuracy than traditional models. (2023-11-08, shares: 3.0)

Investor Risk Appetite and High-Beta Stock Valuation Analysis: The study reveals a pattern in high-beta stock returns around macroeconomic announcements, indicating that investor risk appetite significantly influences these returns. (2023-11-04, shares: 3.0)

Sector Portfolio HRP: Performance and Risk Metrics: A diversified portfolio strategy, Sector Portfolio HRP, outperforms the MSCI All Country World Index in annualized return and risk evaluation from 1996 to 2022, a study shows. (2023-11-03, shares: 4.0)

Identifying Dominance Regimes in the Euro Area with Machine Learning: Machine learning has identified periods of fiscal dominance in the euro area from 2000 to 2019, including during the financial and sovereign debt crises. (2023-11-03, shares: 2.0)

Corporate Culture and Takeover Vulnerability: Research using machine learning indicates that the threat of hostile takeovers can significantly weaken a company's culture, supporting the managerial myopia hypothesis. (2023-11-07, shares: 3.0)

Recently Updated

Quantitative

Bayesian Data Imputation: Missing Data Filling: The article highlights the role of data imputation in risk management, explaining its use in filling gaps in incomplete data for a better understanding of risk factors. (2023-06-28, shares: 189.0)

Machine Learning Execution Time in Asset Pricing: The research analyzes the execution time of machine learning models in empirical asset pricing, finding that XGBoost is the fastest and most accurate, and that reducing features and time observations can significantly cut execution time. (2023-10-31, shares: 2.0)

Interactions in Asset Pricing: Predictors & Returns: The research suggests that future stock returns can be predicted using machine learning models that consider characteristics and macroeconomic variables, resulting in portfolios that perform better than benchmarks. (2023-07-17, shares: 494.0)

Corporate Bonds: Momentum Spillovers: The article uncovers momentum spillovers in the corporate bond market, proposing a strategy of buying bonds from high-performing peers and selling bonds from low-performing peers, yielding a monthly alpha of 36 basis points. (2023-09-25, shares: 2.0)

Alternate Approach: Regression Parameter Estimation: The article presents a new NAS method for univariate regression problems, comparing it with standard methods and suggesting a generalized approach for calculating the cost function's partial derivatives. (2023-09-01, shares: 2.0)

Financial

Efficient Simulation for Derivative Pricing: The article introduces a new simulation-based method for pricing and managing risk of financial derivatives during rare events, proving to be more efficient, accurate, and flexible than traditional methods. (2022-06-08, shares: 85.0)

Commodity Sectors and Factor Strategies: The study explores the impact of commodity sectors on commodity futures risk premiums, revealing that excluding the precious metal sector from a portfolio increases the Sharpe ratio, suggesting precious metals' role as hedging tools affects commodity performance. (2023-10-13, shares: 3.0)

Optimal Valuation Ratio: Forward Price Ratios: The research criticizes the use of trailing price ratios for predicting stock market returns due to changes in cash flow growth, suggesting the use of forward price ratios scaled by cash flow forecasts for better valuation. (2022-12-02, shares: 2.0)

CDS Theory and Practice: The paper examines Quanto Credit Default Swaps, a financial tool that transfers credit risk with foreign exchange exposure, focusing on its theory, pricing, and use in emerging markets like Brazil. (2023-09-15, shares: 75.0)

Volatility Timing with ETF Options: The study finds that hedge funds' positions in ETF options predict volatility in underlying ETF returns, particularly in nonequity ETFs like fixed income and currency ETFs. (2022-10-19, shares: 2.0)

ETF Closures: Do Nothing?: The research indicates that ETFs often close after positive returns and flows, with these factors predicting closure decisions, and smaller ETFs earning higher daily returns than larger ones with the same investment objective. (2023-01-23, shares: 60.0)

Volatility Transformers: Arbitrage-Free Volatility Surfaces: The paper presents a framework for creating arbitrage-free transformations of an implied volatility surface using optimal transport maps, which can be applied to a broader range of synthetic market data generation applications. (2023-09-05, shares: 2.0)

Common Ownership of Stocks & the Low Volatility Anomaly: The study shows that the low volatility anomaly in stock prices is connected to mutual funds performance evaluation against benchmark indexes, as mutual fund managers' heavy investment in certain stocks leads to higher trade volumes and lower volatility. (2023-04-11, shares: 2.0)

ArXiv ML

Recently Published

IGN: A new generative modeling method is suggested, using an idempotent neural network to project any input into a target data distribution. (2023-11-02, shares: 157)

TMKWF: A novel data augmentation technique is proposed that adjusts the distribution of interpolation coefficients based on data point similarity, enhancing model performance and calibration. (2023-11-02, shares: 11)

CM: UHL: A new algorithm is introduced that can learn high-dimensional halfspaces in d-dimensional space in polynomial time, without needing labels. (2023-11-02, shares: 11)

T A PTMF: TopicGPT, a new framework, is introduced that uses large language models to identify latent topics in a text collection, providing more interpretable topics and user control. (2023-11-02, shares: 9)

UniO4: Unifying RL: Uni-o4 is a novel method that merges offline and online reinforcement learning, enhancing the adaptability of the learning process. (2023-11-06, shares: 5)

Reproducible Parameter Inference: The BayesBag study introduces a technique of applying bagging to Bayesian posteriors to enhance reproducibility and uncertainty quantification in model misspecification. (2023-11-03, shares: 5)

PPI: Efficient Inference: PPI++ is a new approach that utilizes a small labeled dataset and a larger machine-learning predictions dataset to boost computational and statistical efficiency. (2023-11-02, shares: 5)

Historical Trending

CodeFusion: Model for Code Generation: CodeFusion is a new model for generating code that improves on previous models by iteratively cleaning up a complete program, matching top systems in initial accuracy and surpassing them in subsequent accuracy checks. (2023-10-26, shares: 3953)

Transport, Variational Inference, and Diffusions in Generative Modeling: The paper introduces a new framework for sampling and generative modelling based on divergences in path space, leading to the creation of a new score-based annealed flow technique and a regularised iterative proportional fitting objective. (2023-07-03, shares: 81)

Explainable Learning of Small Quantum States: The research examines the learned representation of a generative model trained on two-qubit density matrices, showing a direct correlation with the entanglement measure concurrence, providing insights into machine learning of quantum states. (2023-06-09, shares: 22)

RePec

Finance

High-Frequency Alternative Data for GDP Nowcasts: The study uses credit card data to enhance real-time GDP forecasting in Japan, demonstrating that this data improves early-stage forecasting by accurately capturing consumer spending. (2023-11-08, shares: 15.0)

Performance of U.S. ESG ETFs: The paper analyzes the performance of ESG equity ETFs in the U.S. from 2019 to 2021, revealing that these ETFs, on average, outperform the S&P 500 Index. (2023-11-08, shares: 15.0)

High-Dimensional Portfolio Optimization with Factor Model: The article proposes a new portfolio optimization method using a tree-structured portfolio sorting technique, demonstrating that this strategy outperforms others in terms of Sharpe ratios, standard deviation, and turnover. (2023-11-08, shares: 15.0)

Time-Variation in Effects on Portfolio Flows: The research examines the relative significance of push and pull factors for portfolio flows during financial crises, finding that the importance of push factors has increased over time, especially for EU countries. (2023-11-08, shares: 14.0)

Dynamic Bond Portfolio Optimization with Stochastic Interest Rate Model: The paper proposes a new framework for dynamic bond portfolio optimization over multiple periods, which outperforms single-period optimization. (2023-11-08, shares: 26.0)

Multiperiod Portfolio Allocation with Volatility Clustering and Non-Normalities: The study finds that considering volatility clustering in dynamic multiperiod portfolio choices reduces the need for hedging. (2023-11-08, shares: 23.0)

Managed ETFs: Performance Evaluation: A study found that actively managed ETFs in the US from 2018 to 2021 did not yield significant above-market returns and their managers lacked superior market timing skills. (2022-07-09, shares: 18.0)

Statistical

Gender Diversity Prediction in Boardrooms with ML: A study uses machine learning to forecast gender diversity in Chinese company boards, with the extreme Gradient Boosting model showing the best performance. (2023-11-08, shares: 22.0)

Bond Excess Returns Explanation with AI: The SHapley Additive exPlanations technique is used in a paper to pinpoint key factors influencing bond excess return predictions made by machine learning models. (2023-11-08, shares: 21.0)

Signal Quality's Role in Stock Market Volatility Prediction: A study finds that high-quality political signals can predict increased stock market volatility. (2023-11-08, shares: 16.0)

Belief-Based Momentum Indicator and Volatility Predictability in China's Equity Market: Research shows a belief-based momentum indicator can predict equity market volatility in China, with the HAR-LCPR model being the most effective. (2023-11-08, shares: 16.0)

Machine Learning

Deep Learning Model for Newsvendor Problem with Textual Review Data: The article talks about a new inventory management framework that uses a deep learning model. This model suggests order quantities based on online reviews and demand data, reducing costs by 28.7% compared to other models. (2023-11-08, shares: 16.0)

Newsvendor Problem: High-Dimensional Data and Mixed-Frequency Method: The first article explores the application of machine learning to improve demand prediction and restocking decisions in newsvendor problems, utilizing complex and varied historical data. (2023-11-08, shares: 27.0)

Risk-Based Certification for Machine Learning with Continuous Auditing: The second article introduces a Continuous Audit-Based Certification for MLOps systems, which uses automated audits to manage certificates, enhancing efficiency and reliability in the machine learning system. (2023-11-08, shares: 14.0)

GitHub

Finance

Time Series Analysis & Interpretable ML: The article explores Time Series Analysis and Interpretable Machine Learning, focusing on Python packages such as Darts, PyCaret, Nixtla, Sktime, MAPIE, and PiML. (2023-08-19, shares: 13.0)

Fixed Income Library for Bond Pricing & Derivatives: The piece reviews a fixed income library for pricing bonds, bond futures, and derivatives, featuring tools for Curveset construction and risk sensitivity calculations. (2023-03-31, shares: 14.0)

GPU-Accelerated Limit Order Book Simulator for Trading: The article introduces JAXLOB, a GPU-accelerated limit order book simulator aimed at improving large scale reinforcement learning for trading. (2022-04-21, shares: 26.0)

Ultimate Time Series Visualization Tool: The article introduces a Time Series Visualization Tool designed to enhance user experience. (2016-03-01, shares: 3702.0)

Legible Deep Learning with Named Tensors in JAX: The piece explores the use of Named Tensors to improve the readability of Deep Learning in JAX. (2023-06-26, shares: 73.0)

Trending

Optimization: The article is a guide to resources for learning and implementing mathematical optimization, including educational materials and software tools. (2023-10-31, shares: 93.0)

Lock-Free: The article provides a collection of resources for understanding and implementing waitfree and lockfree programming techniques. (2016-03-31, shares: 1565.0)

LaTeX Conversion: The article explores pix2tex, a tool that uses Vision Transformer technology to convert equation images into LaTeX code. (2020-12-11, shares: 6218.0)

LinkedIn

Trending

The Fund: A Dagger on Wall Street: The Fund' has received a positive review from The New York Times, being praised as a sharp critique of Wall Street and the use of money for control and humiliation. (2023-11-07, shares: 2.0)

McKinney Joins Posit: Python data scientist Wes McKinney, known for the pandas package, has joined data science tools company Posit. (2023-11-07, shares: 1.0)

New Causal Modeling Framework: A paper by Lars Lorch, Andreas Krause and Bernhard Schölkopf introduces a new way to discuss causality using Stochastic Differential Equations (SDEs). (2023-11-07, shares: 2.0)

ADGM's DLT Framework Goes Live: The Abu Dhabi Global Market's DLT Foundations Framework, aimed at Blockchain Foundations and DAOs, is now live, advancing Abu Dhabi's commitment to the virtual asset and blockchain sectors. (2023-11-07, shares: 2.0)

Paper on trading with concave price impact: A preprint titled Trading with Concave Price Impact and Impact Decay has been submitted to SSRN, addressing statistical arbitrage issues and estimating trading data. (2023-11-07, shares: 2.0)

Korean tech index soars after short selling ban: South Korean tech index records a 12% single-day gain following a ban on short selling by regulators until June 2024. (2023-11-07, shares: 1.0)

Informative

Math Seminar: Mean Fields Games in Finance: A Math Seminar featuring Prof. Charles-Albert Lehalle will be held on November 9th, 2023, focusing on Mean Fields Games for Financial Markets. (2023-11-07, shares: 2.0)

Advancements in Synthetic Data for AI: The development of GenAI models is hindered by a lack of human-generated data, but synthetic data generation is being utilized by companies like IBM and Google DeepMind. (2023-11-07, shares: 1.0)

European Diversification: Rise of Active ETFs: Active ETFs are becoming increasingly popular in Europe, outperforming active mutual funds which are experiencing significant withdrawals. (2023-11-06, shares: 1.0)

Podcasts

Quantitative

Investors and AI's Impact: A CIO call discusses the potential of artificial intelligence for investors, identifying companies that could benefit or be at risk, and how AI could disrupt the asset management industry. (2023-11-02, shares: 4)

AI and Narratives in Investing: Ben Hunt explores the role of narrative archetypes in understanding artificial intelligence, their influence on industries and money management, and their effect on market trends and investment decisions. (2023-11-06, shares: 4)

The Future of Finance: Quantum Solutions: In the QuantSpeak podcast, Dr. Araceli Venegas-Gomez discusses the potential impact of quantum computing on finance, its adoption in various industries, and her shift from aeronautical engineering to quant finance. (2023-11-06, shares: 4)

ESG Critique by Aswath Damodaran: In a podcast, Professor Aswath Damodaran shares his love for teaching, critiques ESG investing, and discusses market efficiency and the future of education. (2023-11-02, shares: 6)

ESG-Oriented Activism by Impactive Capital: Lauren Taylor Wolfe, co-founder of Impactive Capital, talks about the advantages of a collaborative private equity approach for sustainable long-term returns as an ESG-focused activist investor. (2023-11-03, shares: 6)

Related

Becoming a Legend: Lessons from Fischer Black, Peter Carr, and More: The article highlights the common traits of renowned figures like Fischer Black, Peter Carr, Rick Rubin, George Box, Gilbert Strang, and John Nash, focusing on their soft skills and unique contributions. (2023-11-07, shares: 3)

EM Fixed Income: Analyzing Market Developments Impact: Jonny Goulden and Saad Siddiqui discuss the impact of recent market developments on the EM fixed income asset class in a podcast recorded on 03 November 2023. (2023-11-03, shares: 3)

Debt Buying: Economics, Interest Rates, and Volatility: The recent IBKR Podcast discusses economics, interest rates, and volatility with experts Andrew Wilkinson, Jose Torres, Joe Burke, and Steve Sosnick. (2023-11-07, shares: 4)

Federal Reserve Policies: Insights from Tom Luongo: Tom Luongo analyzes the bond market, Federal Reserve policy, the shift from Libor to SOFR, and potential threats to the Federal Reserve and global markets. (2023-11-02, shares: 3)

Energy, Prosperity, and Investments: Conversation with Chris Martenson: Chris Martenson talks about the link between energy and economic prosperity, the global oil market, the Petrodollar system, and his investment strategies, including gold prices and real rates. (2023-11-07, shares: 3)

News

Unlimited Launches Hedge Fund Replication: Unlimited, a firm co-founded by former Bridgewater Associates executive Bob Elliott, is set to launch a tech-based multi-manager hedge fund replication strategy. (2023-11-08, shares: 3)

Korea prohibits short selling pre-elections: Ahead of the general election, South Korean regulators have completely banned hedge funds and short sellers from selling borrowed stocks, as per Bloomberg. (2023-11-06, shares: 3)

Legal Battle over SEC Power Grab: Hedge funds and private equity firms have started a legal fight against the US Securities and Exchange Commission over the implementation of new regulations. (2023-11-06, shares: 5)

Decline in Hedge Funds: SS&C Technologies reports that hedge funds suffered losses in September, with a decrease in capital movement in October. (2023-11-03, shares: 7)

BNY Mellon Seeks Seed Capital for Newtons Multimanager Push: BNY Mellon is looking for seed capital for Newton Investment Management, its London-based boutique, to break into the multi-manager hedge fund sector. (2023-11-06, shares: 5)

Twitter

Quantitative

RL Algorithmic Trading Strategies in Black Swan Regimes: The article reviews a study assessing the performance of different reinforcement learning trading strategies during unpredictable, extreme market events. (2023-11-03, shares: 5)

Skewness Risk Premium Generates High FX Returns: The article highlights a new study suggesting that trading currencies based on their skewness risk premium can yield high returns and Sharpe ratio. (2023-11-07, shares: 1)

Analyst Underreaction Decline and Momentum Strategy Deterioration: The article discusses a study indicating that the effectiveness of a 12-month momentum strategy has decreased due to analysts' improved reaction to news. (2023-11-05, shares: 1)

Return Drivers of Listed and Unlisted Real Estate: The article examines a study by Chin and Povala that investigates the factors influencing the returns of listed and unlisted real estate, noting a correlation with return horizon. (2023-11-05, shares: 1)

Miscellaneous

Large Language Models for Time Series Forecasting: The NeurIPS 2023 paper presents LLMTime, a large language model that predicts time series data by converting numbers into text and managing missing data. (2023-11-04, shares: 1)

Commodity Strategies and Spreads: The episode offers useful knowledge on commodity strategies and spreads. (2023-11-06, shares: 0)

Microsoft's DeepSpeedRLHF for Chat Inference: Microsoft's DeepSpeedRLHF simplifies chat-style inference, allowing the training of OPT13B in 9 hours and OPT30B in 18 hours for less than $300 and $600 respectively. (2023-11-05, shares: 0)

Theseus: Open Source Library for DNLS Optimization: Theseus is Meta's open-source library for DNLS optimization, developed on PyTorch for structured machine learning. (2023-11-04, shares: 0)

LLMTS: Language Models for Time Series: LLM4TS is a large language model for time series that uses fine-tuning, layer normalization tuning, and LoRA. (2023-11-04, shares: 0)

Reddit

Quantitative

Delta Neutral Strategies: The article explains the profit-making process of delta hedging and delta neutral strategies, using AAPL trading as an example. (2023-11-02, shares: 6.0)

Quant in the Energy Industry: The author recounts their experience in risk management for big energy firms, using python modeling, SQL scripting, and gaining industry-specific knowledge. (2023-11-05, shares: 2.0)

Insightful Books on Quantitative Finance: The author is seeking book recommendations that have greatly influenced others in the field of quantitative finance. (2023-11-04, shares: 71.0)

Quant Trader Noncompete AMA: A former quant trader at a major options market-maker is willing to answer questions about their experience while staying anonymous. (2023-11-05, shares: 227.0)

Quantitative Skill in Undergrad School: The fifth article is not provided. (2023-11-02, shares: 74.0)

Rising

Quant Strategy Building: A robotics engineer with experience in equities seeks advice on creating a personal strategy that merges quantitative and fundamental finance methods. (2023-11-05, shares: 19.0)

Timeframes for High and Medium Frequency Trading: An individual requests clarification on the standard timeframes for high-frequency and medium-frequency trading. (2023-11-03, shares: 19.0)

Active Trading in an IRA for Tax Benefits: A trader using Interactive Brokers Group's API for algorithmic trading considers active trading in an Individual Retirement Account for its tax benefits. (2023-11-02, shares: 24.0)

Essential Languages and Skills for a Quant Trader: A Quantitative Trader with knowledge of R, Python, and Matlab seeks advice on which programming skills to further develop. (2023-11-03, shares: 77.0)

Transitioning to a Career as a Quant Researcher: A data scientist with a Ph.D. in mathematics and experience in the insurance industry is contemplating a career change to quantitative research and seeks guidance for the transition. (2023-11-06, shares: 74.0)

Paper with Code

Trending

DeepSpeed Inference: Efficient Transformer Model at Unprecedented Scale: DeepSpeed Inference improves latency and throughput performance in different situations. (2023-11-07, shares: 1071.0)

OpenChat: Advancing Language Models: The article explores the use of both expert and suboptimal data in general SFT training, without favoring any specific labels. (2023-11-05, shares: 1915.0)

AkariAsai SelfRAG: Learning to Retrieve, Generate, and Critique through Self-Reflection: The framework develops a unique language model that adaptively retrieves and reflects on passages using special 'reflection' tokens. (2023-11-04, shares: 439.0)

Rising

Diffusion Models for Reinforcement Learning: The article emphasizes the superiority of diffusion models over previous generative models in terms of sample quality and training stability. (2023-11-04, shares: 35.0)

GPTFathom: LLM Benchmarking for GPT4+: The rapid advancement of LLMs necessitates an immediate need for a comprehensive evaluation system to identify their pros and cons. (2023-11-05, shares: 146.0)

Language Models with Graph Augmentation for Recommendation: The article discusses strategies to address the issues of sparse feedback and poor side information in recommendation systems. (2023-11-04, shares: 54.0)

Comprehensive Survey on LLM Evaluation: The comprehensive review is designed to stimulate more research into assessing LLMs to ensure their ethical development. (2023-11-04, shares: 192.0)