How-toSelected

Claude Haiku 4.5 memory tests show more notes hurt agents

Why we picked it

The practical shift is architectural: treat agent memory as a curated hint file for style, not a database, and push arithmetic or state tracking into deterministic code. Developers building personal assistants or preference-aware agents can apply this immediately by auditing which memory entries actually reduce errors. The 48% versus 7.1% gap also shows self-rewriting memory still lags far behind explicitly telling the agent the rule.

A test series on personal agents using Claude Haiku 4.5 found that memory size has a sweet spot: rule violations dropped from 77% with no memory to 20% at 10 lines of memory, then rose back to about 25% with longer memories.

Body

Numeric tasks are the weak point, as a stated rule for tracking a running spending total failed 44% of the time, while code that maintained the total failed 0%.

Agents that rewrite their own memory from user complaints improved early but stalled, ending near 48% violations versus 7.

1% when every preference was stated explicitly.The recommendation is to keep memory short, use written rules for style preferences, and delegate anything countable or trackable to code.

Original · rohanpaul_ai

More memory does not keep helping personal agents, because relevant notes help up to a point and then extra lines start making the agent miss rules.

A personal agent can't learn every user preference through memory notes, and more notes eventually make it worse, so use code for anything it must count or track and keep memory short.

Written rules work for style, like signing texts with the user's first name. For a running spending total, a stated rule still failed 44% of the time, while code that kept the total failed 0%.

Memory size has a sweet spot. With Claude Haiku 4.5, violations dropped from 77% with no memory to 20% at 10 lines, then rose to about 25% with longer memories.

Agents that rewrite their own memory from user complaints improve early, then stall. The best methods ended near 48% violations, against 7.1% when the agent was simply told every preference.

Industry Coordinates & Source Verification

This report is categorized under AITOP's "How-to" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Claude Haiku 4.5, agents, memory management, prompt engineering.

Source attribution: curated from rohanpaul_ai (@rohanpaul_ai). Cross-source corroboration: 2 independent sources have covered this development (including arXiv cs.AI), confirming industry-wide signal strength.