arXiv Machine Learning August 21, 2026 By Sawan Dasari

When to Retrain: An Empirical Study of Retraining Policies for Streaming ML Under Concept Drift, Budget, and Latency Constraints

Read the original on arXiv Machine Learning →

arXiv:2608. 19488v1 Announce Type: new Abstract: Production machine learning systems degrade under concept drift, yet practitioners have little principled guidance on when to retrain.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv Machine Learning

5d ago

Differential Refresh Policies for Models Trained on Lagging Data Snapshots: From a Single-Age Equivalence Limit to an Optimal Per-Segment Allocation

arXiv:2610.09519v1 Announce Type: new Abstract: Production machine-learning models are derived artifacts of time-bounded training snapshots: a deployed model is a materialized view over a training cu...

By Amit Rajula

arXiv AI

Aug 24

UpgradeBench: A Decision-Centric Benchmark for Upgrading Fine-Tuned LLM Specialists

UpgradeBench is a decision‑centric longitudinal benchmark that evaluates how fine‑tuned language‑model specialists should be handled when new base‑model releases occur. It covers four consecutive Qwen releases, a continuation checkpoint, six tasks, two model sizes, and OLMo checkpoints with known training lineage, and examines whether retraining, adapter transfer, or other recovery strategies improve specialist performance. The benchmark reveals that upgrade gains vary by task and release interval, that direct adapter copying is sensitive to pretraining distance, and that teacher relabeling can recover specialists without new annotations. "whyItMatters":"The study provides actionable insights into the cost‑effective management of specialist models across model releases, showing how to balance retraining effort with performance gains."

By Ye Chen, Weining Zhang

llms fine-tuning benchmarks data-engineering

arXiv Machine Learning

Aug 5

When RL Meets Adaptive Speculative Training: A Unified Training-Serving System

arXiv:2602. 06932v5 Announce Type: replace Abstract: Speculative decoding can significantly accelerate LLM serving, yet most deployments today disentangle speculator training from serving, treating speculator training as a standalone offline modeling problem.

By Junxiong Wang, Fengxiang Bie, Jisen Li, Zhongzhu Zhou, Zelei Shao, Yubo Wang, Yinghui Liu, Qingyang Wu, Avner May, Sri Yanamandra, Ce Zhang, Tri Dao, Percy Liang, Ben Athiwaratkun, Shuaiwen Leon Song, Chenfeng Xu, Xiaoxia Wu

llms reinforcement-learning

arXiv AI

Oct 1

Pseudo-Label-Triggered Retraining from Forecast Errors for Online Time Series Forecasting

arXiv:2609.39789v1 Announce Type: cross Abstract: Real-world time series forecasting systems operate under non-stationary data streams, where forecasting performance may degrade over time. Although r...

By Yeryeong Kwak, Yoo-Min Jung, Jonghun Park

benchmarks

arXiv AI

Sep 1

Towards Stream Learning on Embedded Systems: Benchmarking the Memory Consumption of Stream Learning Methods

arXiv:2608.30923v1 Announce Type: cross Abstract: Stream learning is commonly evaluated through predictive performance and adaptation to concept drift. However, sustained operation of a stream learne...

By Sebastian Buschj\"ager, Nuwan Gunasekara, Heitor Murilo Gomes

benchmarks

arXiv AI

Sep 30

Where Does Staleness Accumulate? Pool Aware Effective Staleness Control for Asynchronous RL in LLM Post-Training

arXiv:2609.36830v1 Announce Type: new Abstract: Fully asynchronous reinforcement learning (RL) improves resource utilization in large language model post-training by overlapping rollout generation wi...

By Chenliang Li, Neiwen Ling, Zijun Wei, Alfredo Garcia

llms reinforcement-learning benchmarks

One email a morning, machine-written

Related stories

Differential Refresh Policies for Models Trained on Lagging Data Snapshots: From a Single-Age Equivalence Limit to an Optimal Per-Segment Allocation

UpgradeBench: A Decision-Centric Benchmark for Upgrading Fine-Tuned LLM Specialists

When RL Meets Adaptive Speculative Training: A Unified Training-Serving System

Pseudo-Label-Triggered Retraining from Forecast Errors for Online Time Series Forecasting

Towards Stream Learning on Embedded Systems: Benchmarking the Memory Consumption of Stream Learning Methods

Where Does Staleness Accumulate? Pool Aware Effective Staleness Control for Asynchronous RL in LLM Post-Training