Chronofy: A Temporal-Logical Decay Architecture for Information Validity in Time-Aware Retrieval-Augmented Generation

2026-07-17

AI Summary

This work addresses the critical limitation of existing retrieval-augmented generation (RAG) systems in neglecting the temporal validity of knowledge, which often leads to temporal hallucinations. To mitigate this, the authors propose Chronofy, the first three-tier neuro-symbolic architecture that explicitly models time decay throughout the entire RAG pipeline. Chronofy incorporates a structured temporal subspace within Matryoshka embeddings at the representation layer, employs a learnable exponential decay function grounded in Bayesian decision theory at the retrieval layer, and enforces confidence constraints during reasoning via signal temporal logic (STL) robustness combined with the possibilistic weakest-link principle. Experimental results demonstrate that Chronofy significantly improves retrieval accuracy on temporal knowledge graph prediction, TimE question answering, and domain-sensitivity tasks, effectively suppresses the generation of outdated information, and triggers data reacquisition when temporal context is insufficient.

Abstract

Retrieval-Augmented Generation (RAG) systems retrieve and integrate external knowledge to ground large language model (LLM) outputs. However, current RAG architectures treat all retrieved facts as equally valid regardless of temporal provenance, leading to temporal hallucination, where plausible but obsolete facts corrupt the output. A clinical lab reading from yesterday is actionable; the same reading from six months ago is noise. We present Chronofy, a three-layer neuro-symbolic framework implementing the Temporal-Logical Decay Architecture (TLDA) that embeds temporal validity directly into the representation, retrieval, and reasoning layers of RAG systems. Layer 1 reserves a dedicated temporal subspace within Matryoshka embeddings to make fact age structurally irremovable from the representation. Layer 2 integrates learnable exponential decay functions into graph-based retrieval, where the decay coefficient $β_j$ is grounded in Bayesian decision theory as an approximation of twice the latent process mean-reversion rate. Layer 3 applies Signal Temporal Logic (STL) robustness functions to evaluate the temporal validity of retrieved knowledge, not LLM output confidence, and enforces the possibilistic weakest-link principle to bound output confidence by the most decayed evidence in the reasoning chain. We evaluate Chronofy on temporal knowledge graph forecasting benchmarks, the TimE temporal QA benchmark, and a domain-specific sensitivity analysis, demonstrating that explicit temporal decay modeling improves retrieval precision, reduces temporal hallucination, and enables principled data re-acquisition triggers when temporal context is insufficient.

Problem

• temporal hallucination
• information validity
• time-aware retrieval
• retrieval-augmented generation
• temporal decay

Innovation

• Temporal-Logical Decay Architecture
• Retrieval-Augmented Generation
• Signal Temporal Logic
• Matryoshka Embeddings
• Temporal Hallucination