Hallucination Mitigation with Agentic AI, Nested Learning, and AI Sustainability via Semantic Caching
Quick Answer
This paper presents a Nested Learning architecture with Continuum Memory Systems to mitigate hallucinations in LLMs, achieving a Total Hallucination Score reduction of 31.3% to 35.9% across five configurations.
Quick Take
Semantic caching resulted in a 47.3% hit rate, lowering invocations and operational costs, while enhancing factual reliability and auditability without retraining models.
Key Points
- Three-stage agentic pipeline evaluated using five key performance indicators.
- Semantic caching achieved 440 hits out of 930 calls, reducing energy footprint.
- ExtremeObservability configuration yielded the most negative Total Hallucination Score of -0.0709.
- Asymmetric design with high-stochasticity generator improved hallucination mitigation.
- Findings suggest operational efficiency can be enhanced without model retraining.
Paper Resources
Article Content
From source RSS / original summaryarXiv:2605. 29055v1 Announce Type: new Abstract: Hallucination remains a major reliability barrier for production systems, particularly in pipelines where unsupported claims can propagate unchecked across stages. This paper adapts a HOPE-inspired Nested Learning architecture with Continuum Memory Systems (CMS) and semantic similarity caching to a hybrid benchmark of 310 prompts combining 217 epistemic-uncertainty prompts and 93 fabrication-induction stress-test prompts.
A three-stage agentic pipeline orchestrated via the Open Floor Protocol (OFP) is evaluated with five KPIs -- FCD (Factual Claim Density), FGR (Factual Grounding References), FDF (Fictional Disclaimer Frequency), ECS (Explicit Contextualization Score), and OSR (Observability Score Ratio) -- aggregated into THS (Total Hallucination Score) across five weighting configurations to study mitigation-observability trade-offs.
FDF, ECS, OSR, and FGR are subtracted as mitigation signals, so that a more negative THS indicates stronger mitigation. The FrontEndAgent is configured as a high-stochasticity generator (temperature = 1. 0) to produce a realistic hallucination baseline, while the SecondLevelReviewer and ThirdLevelReviewer operate as progressive correctors. This asymmetric design yields end-to-end THS reductions of -31. 3% to -35. 9% across five weighting configurations.
Semantic caching achieves 440 cache hits over 930 potential calls (47. 3% hit rate), reducing LLM invocations to 490, lowering energy and CO2e footprint, and making multi-stage review pipelines operationally viable at production scale. ExtremeObservability attains the most negative final THS (-0. 0709), confirming that observability-heavy configurations reinforce rather than compromise mitigation.
These findings suggest that memory-augmented multi-agent designs can jointly improve factual reliability, operational efficiency, and auditability without model retraining.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.