What I'm reading
Papers, books, essays — things worth the time. Filtered by what actually changed how I think.
Memory Caching: RNNs with Growing Memory
Ali Behrouz, Zeman Li, Yuan Deng, Peilin Zhong, Meisam Razaviyayn, Vahab Mirrokni · 2026
Exploring how recurrent networks can dynamically allocate and manage growing memory buffers for long-term dependencies.
Recursive Reasoning Models — Intuitions from Code
TBD · 2026
Understanding how recursive problem-solving strategies in code can inform the design of models that reason hierarchically.
Attention Is All You Need
Vaswani, Shazeer, Parmar, Uszkoreit, Jones, Gomez, Kaiser, Polosukhin · 2017
The architecture that started everything. Read it to understand why everyone keeps talking about Q, K, V. Worth reading twice — once for the math, once for the intuition.
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Wei, Wang, Schuurmans, Bosma, Ichter, Xia, Chi, Le, Zhou · 2022
Simple idea with outsized impact — just show the model how to think step by step. Read this to understand why prompt structure matters more than prompt length.
Keep CALM and Explore: Language Models for Action Generation in Text-based Games
Yao, Rao, Hausknecht, Narasimhan · 2020
The way the AI learns to play text-based games on its own was fascinating. I liked how the paper combines language understanding with reinforcement learning, especially the idea of generating action candidates from human gameplay patterns.
Memory in the Age of AI Agents
Hu, Liu, Yue, Zhang et al. · 2025
Currently discussing in an online ML community. The three-dimensional taxonomy — forms, functions, dynamics — reframes how I think about RAG. Most memory frameworks are footnotes on this structure.
Introduction to Algorithms (CLRS)
Cormen, Leiserson, Rivest, Stein · 2009
I genuinely found this book extremely good for learning algorithms quickly. The explanations and structured approach helped me understand problem solving and logic in a much clearer way.
Chap 3: Feature Driven Development Practices — A Practical Guide to FDD
Stephen Palmer, Mac Felsing · 2002
The practical companion to FDD theory. The design-by-feature and build-by-feature split is the part that actually transfers to real work.
Chapter 23: Adaptive Software Development
Ken Orr · 2001
The primary source on ASD. The emphasis on emergence over planning resonates with how DNA was actually run.
Research Methodology for Beginners
Kitasakorn Locharoenrat · 2013
Read cover to cover for MCA. More useful than expected — the sections on research design directly changed how I approach technical projects.
Adaptive Software Development
TutorialsPoint · 2022
Overview-level reading for ASD. Good primer before going to the Ken Orr chapter.
Crystal Agile Methodology: A Comprehensive Study
Aditi Mahatre · 2020
Crystal's idea of calibrating process weight to project criticality stuck with me — more damage potential = more ceremony. Simple logic.
Feature Driven Development
Jeff De Luca (ref: Brooks 1995, Davis & Meyer 1998) · 2002
FDD's feature-centric view made more practical sense to me than sprint-based planning. Shipping features > completing sprints.
Process Diversity in Software Development
Mikael Lindvall, Ioana Rus · 2002
Coursework reading. Makes the case that no single process fits all projects — obvious in hindsight, but worth having language for.
Selecting a Project Methodology
Alistair Cockburn · 2000
Cockburn's framework for matching methodology to project type. Useful heuristic: the more people, the more ceremony you need.
Story Card Maturity Model: A Process Improvement Framework for Agile Requirement Practices
Chetankumar Patel, Muthu Ramachandran · 2009
Read for MCA. The maturity model framing for something as lightweight as story cards was an interesting contrast.