---
title: "PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2608.04003v1"
permalink: "https://feed.astralor.com/items/itm_6a95ac2c4a8ad5cc00b872a8"
published: "2026-08-04T17:58:05.000Z"
score: 77
channel: "ai"
tags: ["Hermes+","PAST-Bench","Personal AI Agents","Benchmark","Recursive Self-Improvement"]
---

# PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

## AI commentary

> What stands out is that the paper goes beyond headline numbers and uses pathway evidence to expose superficial improvements. For agent product builders, Hermes+'s five interventions are highly concrete, especially the outdated-state replacement step, which directly targets a common pain point. If PAST-Bench gains community adoption, it could become the standard litmus test for measuring a personal agent's learning capability.

## At a glance

Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents retain preferences, task histories, tool routines, and learned skills across sessions, making them an ideal setting for studying this capability—yet whether retained experience truly improves agents over time has not been systematically tested. PAST-Bench isolates this question through 26 scenarios and 204 episodes, comparing agents with experience retention on and off across memory, procedural reuse, information gathering, and update capabilities. Across seven base models and four agent frameworks, improvement is real but uneven. Guided by these findings, Hermes+ introduces five targeted interventions and raises average gains, strongest on tasks requiring outdated state to be replaced, though effects remain capability- and model-dependent.

## Key points

- PAST-Bench features 26 scenarios and 204 episodes with matched on/off retention conditions to isolate recursive self-improvement.
- It evaluates memory, procedural reuse, information gathering, and update across seven base models and four agent frameworks.
- Improvement is real but uneven, and agents with similar headline gains can differ in whether those gains follow the intended pathway.
- Guided by findings, Hermes+ adds five targeted interventions, raising average gains and strongest on outdated-state-replacement tasks.

## Full text

Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents offer a concrete setting for studying this capability because they retain preferences, task histories, tool routines, and learned skills across sessions. Yet whether retained experience actually improves them over time has not been systematically tested. We introduce PAST-Bench, a benchmark designed to isolate this question. Each agent runs through ordered sequences of fresh-session tasks under matched conditions that turn retained experience on and off. It spans 26 scenarios and 204 episodes across memory, procedural reuse, information gathering, and update. We report both later-task gains and whether those gains follow the intended save, retrieve, and update pathway. Across seven base models and four agent frameworks, improvement is real but uneven across capabilities. Agents with the same headline gain can differ markedly in whether that gain is supported by evidence of the intended pathway. Guided by these findings, we develop Hermes+, which extends Hermes with five targeted interventions across stages of the agent loop. Hermes+ raises the average gain from retained experience and provides clearer pathway evidence, with its strongest improvement on tasks requiring outdated state to be replaced, although the effect remains capability- and model-dependent. Together, PAST-Bench and Hermes+ provide an evaluation and diagnostic foundation for studying how persistent agents can progress from retaining experience to systematically improving through it. Code: https://github.com/Gen-Verse/PAST-Bench

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_6a95ac2c4a8ad5cc00b872a8