Session-Isolated Unfalsified Adaptive Control for Multi-Turn Intent Containment in Autonomous Agent Architectures
- 1Bangsaen AI Labs
Abstract
The security of multi-turn autonomous agent architectures is fundamentally compromised by context-steering vulnerabilities, specifically Salami attacks, where adversarial intent is incrementally dispersed across sequential interaction turns. Conventional probabilistic filters and static language guardrails fail under such trajectory-based manipulation due to the lack of dynamic state memory and the presence of model-vs-payload space mismatch.
To resolve this challenge at its mathematical core, we formulate multi-turn intent containment as a data-driven, state-space control problem governed by Unfalsified Control Theory and Koopman Operator dynamics. By lifting discrete token interaction streams into an infinite-dimensional Hilbert space, the nonlinear evolution of multi-turn dialogue trajectories is linearized into observable subspace projections. We define a dual-subspace energy framework that continuously evaluates the instantaneous directional momentum of user intent relative to an immutable, zero-calibration energy density hyperplane boundary.
To eliminate subspace saturation and contextual information decay during extended multi-turn sessions, we introduce a session-isolated adaptive dual-loop learner augmented with a sliding-window forgetting operator. This mechanism guarantees bounded memory dissipation while preserving short-term intent trajectory momentum. Upon falsification—defined strictly by a negative Lyapunov-like energy density derivative—the system executes an immediate, deterministic state-transition into a closed-loop containment sandbox, accompanied by a zero-remanence memory sanitization protocol that destroys volatile trajectory information.
By operating entirely as a deterministic, Python-native linear algebra transformation, the proposed architecture provides mathematically guaranteed, real-time intent containment for both digital API gateways and physical kinetic actuator interfaces without reliance on offline retraining, empirical threshold tuning, or hyperparameter optimization.
Related
- Deterministic Execution Fencing and Write-Ahead Containment for Autonomous Systems
- Agentic AI: A Comprehensive Review of Autonomous Agents, Architectures, and Democratized Implementation for Easy Agent Creation
- Exact Pharmacokinetic Washout Bounds, Cycle Conservation Laws, and Model-Free Resistance Observers in Adaptive Cancer Containment
- Optimal Zonotope Containment via Adaptive Gaussian Sampling
- Autonomous AI-Driven Nanoscale Spatial Mapping Reveals Novel Targets and Ternary Architectures in 5xFAD Alzheimer's Disease Model
BibTeX
@misc{wonghong2026sessionisolated,
title = {{Session-Isolated Unfalsified Adaptive Control for Multi-Turn Intent Containment in Autonomous Agent Architectures}},
author = {Tanet Wonghong},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23555956},
url = {https://zenodo.org/records/23555956},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Session-Isolated Unfalsified Adaptive Control for Multi-Turn Intent Containment in Autonomous Agent Architectures
AU - Tanet Wonghong
PY - 2026
DA - 2026/10/22
PB - Zenodo
DO - 10.5281/zenodo.23555956
UR - https://zenodo.org/records/23555956
AB - The security of multi-turn autonomous agent architectures is fundamentally compromised by context-steering vulnerabilities, specifically Salami attacks, where adversarial intent is incrementally dispersed across sequential interaction turns. Conventional probabilistic filters and static language guardrails fail under such trajectory-based manipulation due to the lack of dynamic state memory and the presence of model-vs-payload space mismatch. To resolve this challenge at its mathematical core, we formulate multi-turn intent containment as a data-driven, state-space control problem governed by Unfalsified Control Theory and Koopman Operator dynamics. By lifting discrete token interaction streams into an infinite-dimensional Hilbert space, the nonlinear evolution of multi-turn dialogue trajectories is linearized into observable subspace projections. We define a dual-subspace energy framework that continuously evaluates the instantaneous directional momentum of user intent relative to an immutable, zero-calibration energy density hyperplane boundary. To eliminate subspace saturation and contextual information decay during extended multi-turn sessions, we introduce a session-isolated adaptive dual-loop learner augmented with a sliding-window forgetting operator. This mechanism guarantees bounded memory dissipation while preserving short-term intent trajectory momentum. Upon falsification—defined strictly by a negative Lyapunov-like energy density derivative—the system executes an immediate, deterministic state-transition into a closed-loop containment sandbox, accompanied by a zero-remanence memory sanitization protocol that destroys volatile trajectory information. By operating entirely as a deterministic, Python-native linear algebra transformation, the proposed architecture provides mathematically guaranteed, real-time intent containment for both digital API gateways and physical kinetic actuator interfaces without reliance on offline retraining, empirical threshold tuning, or hyperparameter optimization.
ER -