Start Here
A short map of the main themes on this site.
This site is about the practical side of ML and AI systems: how models, agents, training pipelines, evaluations, and product interfaces behave when they leave the paper and enter a real engineering loop.
The fastest entry point is to pick a path below, then follow the linked topic hub when you want the full archive.
Reading Paths
LLM Agents
Tool use, agent runtime design, evaluation, context, and production patterns for systems that act across tools and environments.
- Improving LLM Internationalization: Bridging the Gap in Tool Use and Agency
A practical multilingual agent playbook, updated with task-specific evidence from MASSIVE-Agents, GAIA-v2-LILT, and SEATauBench, with explicit limits on what transfers to current frontier models.
- What Makes an Agent Swarm Work? Simple Rules, Information Boundaries, and the Value of Verification
A theory-and-experiment study of agent swarm organization: operational roles, marginal-contribution incentives, selective information sharing, and why verification remains the limiting assumption.
- Do Verifier Errors Grow Superlinearly with Horizon? A Six-Stage Experiment
A preregistered 224-artifact study confirms a 66% hybrid-verifier improvement, does not support the universal superlinear headline, and finds that task structure can reverse the apparent horizon effect.
Evaluation
Practical approaches to measuring model and agent capability with deterministic checks, rubrics, trajectories, and verifiable outcomes.
- From GRPO Outcome Rewards to Token-Level Advantage
A practical framework for turning GRPO-style sequence rewards into token-level advantages, including GAE-style estimators, credit assignment routes, and multi-reward training design.
- When Does Model Souping Work for LLMs?
A practical guide to when LLM weight averaging and model merging work, why they fail, how methods such as Task Arithmetic, TIES, DARE, and LoRA merging differ, and how to evaluate a merge before deployment.
- Do LLM Agents Work Equally Well Across Languages?
Agent performance differs across languages, tasks, and localization settings. A visual review of 30 languages and 16 evidence sources identifies where Arabic, Thai, Japanese, and Hindi need targeted evaluation.
Post-Training
SFT, RLHF, preference optimization, instruction following, reasoning traces, and data pipelines for shaping model behavior after pretraining.
- From GRPO Outcome Rewards to Token-Level Advantage
A practical framework for turning GRPO-style sequence rewards into token-level advantages, including GAE-style estimators, credit assignment routes, and multi-reward training design.
- When Does Model Souping Work for LLMs?
A practical guide to when LLM weight averaging and model merging work, why they fail, how methods such as Task Arithmetic, TIES, DARE, and LoRA merging differ, and how to evaluate a merge before deployment.
- What Makes an Agent Swarm Work? Simple Rules, Information Boundaries, and the Value of Verification
A theory-and-experiment study of agent swarm organization: operational roles, marginal-contribution incentives, selective information sharing, and why verification remains the limiting assumption.
RLHF and Preference Optimization
Engineering notes and research synthesis on PPO, DPO, GRPO, reward modeling, preference data, and model behavior optimization.
- From GRPO Outcome Rewards to Token-Level Advantage
A practical framework for turning GRPO-style sequence rewards into token-level advantages, including GAE-style estimators, credit assignment routes, and multi-reward training design.
- What Makes an Agent Swarm Work? Simple Rules, Information Boundaries, and the Value of Verification
A theory-and-experiment study of agent swarm organization: operational roles, marginal-contribution incentives, selective information sharing, and why verification remains the limiting assumption.
- Do Verifier Errors Grow Superlinearly with Horizon? A Six-Stage Experiment
A preregistered 224-artifact study confirms a 66% hybrid-verifier improvement, does not support the universal superlinear headline, and finds that task structure can reverse the apparent horizon effect.
Generative UI
How AI systems can produce, steer, and execute user interfaces with structured representations and practical product constraints.
- Ad Formats in LLM Products: What's Live vs. What's Research
A survey of advertising formats in LLM products—separating what's deployed in production from what remains in research.
- Generative UI Doesn't Move the Needle—Steering Does
After shipping multiple generative UI features, I've concluded that the sophistication of AI-generated interfaces often doesn't translate to user benefit—but steering does.
- UI Representation and Action Execution for Generative UI
Exploring structured UI representation using JSON Schema, and how to implement action handlers for generative UI systems.