Metraipolis › archive › dc_e8ffce43b2a0

CausalCache: Conditional High-Fidelity Restoration for Long-Horizon GUI Agents — arXiv:2608.22577

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.22577
License (as archived)
CC0-1.0
Archived
2026-10-18T07:41:31.177Z
SHA-256 of the archived text
240e3ce3fd5e2606634b3a281bbb66fc2d5daa6343b31df1e9cd44a17c099e4a
Stable URL of this page
https://metraipolis.com/documento/dc_e8ffce43b2a0

How to cite

CausalCache: Conditional High-Fidelity Restoration for Long-Horizon GUI Agents — arXiv:2608.22577 — arXiv (en), https://arxiv.org/abs/2608.22577 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e8ffce43b2a0 (Metraipolis archive dc_e8ffce43b2a0, sha256 240e3ce3fd5e2606634b3a281bbb66fc2d5daa6343b31df1e9cd44a17c099e4a)
BibTeX
@misc{dce8ffce43b2a0,
  title = {CausalCache: Conditional High-Fidelity Restoration for Long-Horizon GUI Agents — arXiv:2608.22577},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.22577},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e8ffce43b2a0; sha256 240e3ce3fd5e2606634b3a281bbb66fc2d5daa6343b31df1e9cd44a17c099e4a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.22577. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-horizon GUI agents can retain complete action histories as compact text, but only a few historical screenshots fit in active context. We formulate this as budgeted fidelity restoration: every event remains summarized, while a fixed budget $B$ determines which events regain their archived screenshots. Recent-$B$ assigns all visual slots to the latest events. CausalCache instead scores the complete history and swaps in an older event only when its predicted utility exceeds that of a recent event. A history-gated key/value adapter modifies only restored history-image tokens and is exactly bypassed when no history image is active, preserving current-screen processing. The adapter and selector are trained with matched-budget interventions on desktop trajectories and evaluated zero-shot on mobile. On OSWorld-Verified, activating historical screenshots improves success by about $13$ percentage points over summary-only memory. Under the official $15$-step limit, CausalCache and Recent-$4$ are statistically indistinguishable; in a $30$-step diagnostic, CausalCache achieves $46.7\%$ success versus $42.4\%$ ($+4.3$ points). Zero-shot on $117$ MobileWorld tasks, CausalCache improves over Recent-$4$ from $30.2\%$ to $36.8\%$. The gain is concentrated on a pre-defined cross-app memory-candidate split ($30.6\%$ vs. $19.4\%$, $+11.2$ points), while single-app controls show no detectable difference ($43.6\%$ vs. $42.4\%$). These results show that selecting which past events regain pixels is more effective than spending a fixed visual budget entirely on recency.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.22577 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T07:37:52.729Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).