Series

42 articles

All Series Insight Tutorial
Part 7. TANREN — Atari Bowling, the Lost Campaign That Built the Entry Gate: Weighing a Fight Before Running It Series
TANRENEvolutionary SearchFunSearchAtariReinforcement LearningLLMPython

Part 7. TANREN — Atari Bowling, the Lost Campaign That Built the Entry Gate: Weighing a Fight Before Running It

On Atari Bowling, TANREN beat R2D2, the human benchmark, and DQN, then hit a ceiling at 74% of the strongest RL. This is the record of establishing — by measurement — that the ceiling was structural to the class of readable reactive code, and of how that loss produced the entry gate that now guards every experiment.

Read more →
Part 5. TANREN — Discovering Breakout's Tunnel Strategy on Its Own, and Measuring Exactly Why the Rest Is Out of Reach Series
TANRENEvolutionary SearchFunSearchAtariReinforcement LearningLLMPython

Part 5. TANREN — Discovering Breakout's Tunnel Strategy on Its Own, and Measuring Exactly Why the Rest Is Out of Reach

In Breakout, evolution discovered the classic tunnel strategy without being taught it and reached about 6.1x the human benchmark. Just as important, the reason it could not reach the strongest RL agents was pinned down by measurement, not speculation — drawing the boundary of what a few dozen readable lines of code can do.

Read more →
Part 2. TANREN — 11 Wins to 1 Against Twenty Years of Cache Classics, on the Future of a Production Trace Series
TANRENEvolutionary SearchFunSearchCachingLLMPythonvLLM

Part 2. TANREN — 11 Wins to 1 Against Twenty Years of Cache Classics, on the Future of a Production Trace

Cache eviction — deciding what to drop when a cache is full — has been studied for more than twenty years. On that ground, a 53-line evolved Python function beats ARC, LIRS, and W-TinyLFU on future segments of Twitter's production traces. This post covers how the first attempt lost by overfitting to one time period, the two-window training that fixed it, and the public code anyone can rerun.

Read more →
Part 8. TANREN — The Number I Almost Claimed and Then Withdrew: Atari Freeway and Measurement Honesty Series
TANRENEvolutionary SearchFunSearchAtariReinforcement LearningLLMPython

Part 8. TANREN — The Number I Almost Claimed and Then Withdrew: Atari Freeway and Measurement Honesty

A score of 31.9 looked like a DQN-beating result — until re-measuring under the published side's exact protocol dropped it to 28.6. This is the record of withdrawing the number, repairing the verifier, and re-running the evolution under the correct protocol to reach 31.2: past the human benchmark and DQN, at 96% of Agent57. An article about evaluation honesty.

Read more →
Part 6. TANREN — On Atari Skiing, Deep RL's Hardest Slope, Readable Python Beats Every Published RL Agent Series
TANRENEvolutionary SearchFunSearchAtariReinforcement LearningLLMPython

Part 6. TANREN — On Atari Skiing, Deep RL's Hardest Slope, Readable Python Beats Every Published RL Agent

On Atari Skiing — where Agent57 needed roughly 78 billion training frames to surpass the human benchmark — about 130 lines of readable Python scored −3310.7 in a held-out evaluation, beating every published RL agent and the human benchmark. Includes the entry gate that estimated the odds before running, the techniques evolution added on its own, and the cost contrast: one day and a few dollars.

Read more →
Part 4. TANREN — an Evolved Scheduler Patched into vLLM Halves Tail Latency Under Heavy Load Series
TANRENEvolutionary SearchvLLMLLMSchedulingPythonGB10

Part 4. TANREN — an Evolved Scheduler Patched into vLLM Halves Tail Latency Under Heavy Load

The scheduler — the heart of LLM serving — evolved by TANREN and A/B tested inside a real vLLM. Across 16 real-trace configurations it beats the default FCFS 15 times (p=2.6e-4), cuts mean/p99 TTFT by 35-49% under heavy load, and does no harm when idle. Includes the honest record of going 0/6 on unseen data and six rounds of diagnosis before the win.

Read more →
Part 3. TANREN — Patching the Evolved Cache Policy into a Real vLLM and Measuring It Series
TANRENEvolutionary SearchvLLMLLMCachingPythonGB10

Part 3. TANREN — Patching the Evolved Cache Policy into a Real vLLM and Measuring It

The cache policy that won in simulation is patched into a live vLLM and measured on real hardware. The simulated win vanished at first; after fixing where and what to measure, tail latency (p99 TTFT) came out 13-16% lower under memory pressure. The prototype's costs and the conditions where it does not help are reported as measured.

Read more →
Part 1. TANREN — Atari Pong 21–0, and the Kill-Shot the Loop Found on Its Own Series
TANRENEvolutionary SearchFunSearchAtariReinforcement LearningLLMPython

Part 1. TANREN — Atari Pong 21–0, and the Kill-Shot the Loop Found on Its Own

How an evolutionary loop — a cheap LLM writing code, scored by a deterministic verifier — reached the theoretical ceiling on Atari Pong (21–0 across 40 games) from nothing but the 128-byte RAM, and discovered a single winning shot without being taught it. Includes the honest protocol and the diagnostic loop that made it work for a few dollars.

Read more →
Part 2. Measuring KOI on a Text-to-SQL agent — picking a cost-effective model without dropping quality Series
KOIText-to-SQLLLMAI Cost OptimizationOpenRouterNishikiAI Agent

Part 2. Measuring KOI on a Text-to-SQL agent — picking a cost-effective model without dropping quality

Using a Text-to-SQL agent — you ask in plain language, the AI writes SQL and queries an internal database — I measured five models by KOI (KPI ÷ cost). With a quality floor in place, I show how to pick the most cost-effective model, with real numbers and the full procedure.

Read more →