Context Windows Forget What Matters — I Built a Usage-Reinforced Decay Engine for AI Agent Memory
Emmimal P Alexander Jul 24, 2026 23 min read Share Image by the author, generated with ChatGPT (DALL·E) TL;DR: Most long-running AI agent memory systems evict old context using a sliding window: if something hasn’t been touched in N turns, it’s gone. This treats a foundational fact stated once on turn 1 and a throwaway debug log from turn 40 as identical — whichever is older loses, no matter how many times either one has actually been used. I built a memory engine that scores retention using the Ebbinghaus forgetting curve, where every recall reinforces an item’s stability and pushes its eviction horizon out non-linearly.
- ▪Emmimal P Alexander Jul 24, 2026 23 min read Share Image by the author, generated with ChatGPT (DALL·E) TL;DR: Most long-running AI agent memory systems evict old context using a sliding window: if something hasn’t been touched in N turns,
- ▪This treats a foundational fact stated once on turn 1 and a throwaway debug log from turn 40 as identical — whichever is older loses, no matter how many times either one has actually been used.
- ▪I built a memory engine that scores retention using the Ebbinghaus forgetting curve, where every recall reinforces an item’s stability and pushes its eviction horizon out non-linearly.
Towards Data Science files mainly under ai. We currently carry 104 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | Towards Data Science |
| Canonical URL | https://towardsdatascience.com/context-windows-forget-what-matters-i-used-a-140-year-old-psychology-paper-to-fix-ai-memory/ |
| Publication time | Fri, 24 Jul 2026 12:00:00 +0000 |
| Retrieval time | 2026-07-24T12:22:43.243Z |
| Last seen | 2026-07-24T12:22:43.243Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | DO8wLkXlb3eZ |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
Large Language Models Context Windows Forget What Matters — I Built a Usage-Reinforced Decay Engine for AI Agent Memory How to build a usage-reinforced memory decay engine for AI agent memory that outperforms naive recency-window pruning on long-running, multi-session tasks — with a deterministic, zero-dependency Python implementation you can verify yourself. Emmimal P Alexander Jul 24, 2026 23 min read Share Image by the author, generated with ChatGPT (DALL·E) TL;DR: Most long-running AI agent memory systems evict old context using a sliding window: if something hasn’t been touched in N turns, it’s gone.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Towards Data Science.
Discussion
0 commentsMore from Towards Data Science
