# Risk-Constrained Freshness-Aware Semantic Caching for Open-Web Retrieval-Augmented LLMs

arXiv: https://arxiv.org/abs/2607.04281 · HTML: https://mlsanity.com/paper/2607.04281/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Generation and context integration, one of 15 directions in Retrieval-augmented generation.
- Builds on 12 of its 23 references that are in the map.
- First posted on arXiv 5 Jul 2026.

## Placed in

- Retrieval-augmented generation / [Generation and context integration](https://mlsanity.com/rag/generation-and-context-integration/)

## Builds on (most cited first, 12 of 12)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022)
- [Time-Aware Language Models as Temporal Knowledge Bases](https://mlsanity.com/paper/2106.15110/) (2021)
- [RealTime QA: What's the Answer Right Now?](https://mlsanity.com/paper/2207.13332/) (2022)
- [CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusion](https://mlsanity.com/paper/2405.16444/) (2024)
- [SealQA: Raising the Bar for Reasoning in Search-Augmented Language Models](https://mlsanity.com/paper/2506.01062/) (2025)
- [Category-Aware Semantic Caching for Heterogeneous LLM Workloads](https://mlsanity.com/paper/2510.26835/) (2025)

## Built on by (most cited first, 2 of 2)

- [Semantics Delivery Network: Rethinking Web Retrieval Infrastructure for LLM Agents](https://mlsanity.com/paper/2609.22486/) (2026)
- [PACE: Perceived-Latency-Aware Cascading Service Routing and Filler Control for QoE-Efficient Retrieval-Augmented Dialogue Serving](https://mlsanity.com/paper/2609.10372/) (2026)

## Abstract (arXiv)

Semantic caching reduces the latency and cost of retrieval-augmented generation (RAG) by serving cached answers to semantically similar queries, but most existing methods do not model the time-varying freshness of open-web evidence. We present FreshCache, a three-tier semantic cache that treats cache reuse as a risk-constrained temporal inference problem: before approving a cache hit, FreshCache estimates the probability that the cached result is stale using a fitted exponential decay model enhanced by a learned MLP, and approves reuse only when that probability falls below a per-tier error budget across answers (epsilon = 0.10), URL lists (epsilon = 0.20), and page content (epsilon = 0.35). This allows the system to degrade gracefully as entries age rather than forcing a binary choice between a stale hit and a full pipeline execution. We introduce FreshCache-Bench, a benchmark of 8,072 base queries across five freshness classes with ground truth staleness labels drawn from real web snapshots at 1, 12, 24 hours, and 7 days after a baseline crawl, expanded to 31,201 queries via paraphrase generation. At the 24-hour evaluation window, FreshCache_MLP achieves 97% search API savings at 0.1% hash-based stale error, and an LLM-judge evaluation on 396 confirmed change pairs shows that only 34.3% of detected content changes actually affect answer correctness, placing true answer-affecting stale error at approximately 0.034%. The rule-based FreshCache achieves 98% search savings at 3.3% stale error under a temporal holdout calibration, outperforming SemanticTTL (14.9% stale, 72% saved), vCache (7.2% stale, 47% saved), and SCALM (5.2% stale, 96% saved). Ablations show the temporal risk gate accounts for an 11.6 point reduction in stale error over similarity-only reuse, and the learned MLP reduces stale error a further 3.2 points over the rule-based model.