Computation of Strong Solutions to Stochastic Variational Inequalities
Builds a computational framework for monotone stochastic variational inequalities, connecting optimization algorithms with stochastic modeling.
Open arXiv ↗Reading notes · papers · GitHub projects
Learn With Me
A living page for papers, repositories, and ideas I am following in mathematics, machine learning, and AI agents. I use this as a public learning trail: what looks interesting, why it matters, and what I want to understand next.
Last curated: 2026-09-06
01
Read actively
For each paper or repo, I focus on the core problem, assumptions, method, and what I can reuse.
02
Connect ideas
I group resources by themes so math, ML, and agent systems can cross-pollinate each other.
03
Build small
When a resource looks useful, the next step is a small experiment, reproduction, or note.
Probability, dynamics, optimization, and mathematical foundations
Mathematics
Papers and repositories that connect rigorous mathematical ideas with modeling, simulation, and computational tools.
Thermodynamic Concentration Inequalities: Controlling Uncertainty in Finite-Time and Small-Sample Thermodynamic Inference
Derives nonasymptotic concentration bounds for currents of ergodic diffusions, a useful bridge between probability and uncertainty-aware physical inference.
Open arXiv ↗statsmodels/statsmodels
A mature Python library for statistical modeling and econometrics, useful for learning how rigorous statistical methods are exposed through production scientific software.
Open GitHub ↗Greedy sampling designs via reduced basis methods: optimal recovery in the uniform norm
Connects RKHS sampling, reduced-basis ideas, and uniform-norm recovery, offering a rigorous lens on where to sample complex functions.
Open arXiv ↗stdlib-js/stdlib
A substantial numerical and scientific standard library for JavaScript/TypeScript, useful for studying how mathematical tooling is organized in production.
Open GitHub ↗oscar-system/Oscar.jl
A comprehensive Julia-based computer algebra system for algebra, geometry, and number theory, useful for seeing modern mathematical software design in practice.
Open GitHub ↗Deep learning, evaluation, optimization, and applied ML systems
Machine Learning
A running list of ML papers and codebases I am reading to improve model-building, evaluation, and deployment intuition.
Compile by Training: Turning Natural-Language Specifications into Local Neural Functions
Explores replacing repeated remote LLM calls with trained local neural functions, a practical idea for lowering cost, latency, and provider dependence.
Open arXiv ↗Clean Engineering, Unstable Measurement: A Preregistered Reliability Failure of Black-Box LLM Observers on Shared Endpoints
Tests black-box LLM judges as measurement instruments and highlights reliability risks that matter for evaluation pipelines and leaderboards.
Open arXiv ↗Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning
Challenges the assumption that readable chain-of-thought faithfully explains model decisions, sharpening how to think about interpretability.
Open arXiv ↗scverse/scvi-tools
A widely used toolkit for deep probabilistic analysis of single-cell and spatial omics data, showing how modern ML methods become reusable scientific workflows.
Open GitHub ↗vllm-project/vllm
A high-throughput LLM inference engine worth tracking for real-world serving, batching, and memory-management patterns.
Open GitHub ↗sgl-project/sglang
A fast serving framework for large language and multimodal models, complementing vLLM and TensorRT-LLM with another production-oriented inference stack to study.
Open GitHub ↗Tool use, memory, autonomy, benchmarks, and agent infrastructure
AI Agents
Resources for understanding how AI agents plan, use tools, evaluate themselves, and operate as long-running systems.
A Case Study on Emergent Cheating and Whistleblowing in Autonomous Research Swarms
Studies how multi-agent research systems can develop coordination failures, making it relevant to safety, incentives, and monitoring for agent teams.
Open arXiv ↗SENTINEL-RL: Offloading Topological Reasoning from LLM Agents in the Security Operations Center
Pairs LLM security agents with specialized topological reasoning, showing a concrete pattern for tool-augmented agents in complex environments.
Open arXiv ↗callstack/agent-device
Provides mobile app automation and verification tools for AI coding agents, a practical bridge between agent workflows and real device testing.
Open GitHub ↗PatchBench: Evaluating AI Agents for Vulnerability Patching
Introduces a benchmark for vulnerability-patching agents that goes beyond simple proof-of-concept crash checks toward more realistic evaluation.
Open arXiv ↗agno-agi/agno
A widely used framework for building and managing agent platforms, useful for studying memory, tools, workflows, and production agent infrastructure.
Open GitHub ↗promptfoo/promptfoo
A practical evaluation and red-teaming toolkit for prompts, agents, and RAG systems, helpful for learning how agent behavior is tested before deployment.
Open GitHub ↗