APEX: New Framework Improves Prompt Optimization for Large Language Models Through Dynamic Data Selection
Researchers have introduced APEX, a framework that optimizes both prompts and data usage simultaneously for large language models. Unlike existing evolutionary approaches that treat evaluation datasets as static benchmarks, APEX dynamically categorizes data into Easy, Hard, and Mixed tiers to focus compute on the most informative examples. The system achieved average accuracy improvements of 11.2% on Gemini 2.5 Flash and 6.8% on Gemma 3 27B under a fixed budget of 5,000 evaluation calls.
A team of researchers has proposed APEX (Automatic Prompt Engineering eXpert), a new framework designed to address a key inefficiency in current automated prompt optimization methods. Existing evolutionary algorithm-based approaches treat the development dataset as a static resource, spending compute budget on data points that provide little signal about prompt quality. APEX counters this by dynamically stratifying the dataset into Easy, Hard, and Mixed tiers based on the optimization history, then prioritizing the Mixed tier where model performance is inconsistent. Within that tier, the framework identifies two high-leverage subsets: an 'addressable frontier' for generating informative prompt mutations, and a 'rank-sensitive frontier' for distinguishing between candidate prompts. Evaluated on three benchmarks—IFBench, SimpleQA Verified, and FACTS Grounding—APEX outperformed baseline prompts by an average of 11.2% on Gemini 2.5 Flash and 6.8% on Gemma 3 27B, all within a fixed evaluation budget, suggesting that data-centric strategies can meaningfully improve prompt optimization efficiency.
What's missing
The study evaluates APEX only on two specific models (Gemini 2.5 Flash and Gemma 3 27B); generalizability to other LLM families is untested.
What different sources said
- arXiv cs.AICenter
APEX: Automated Prompt Engineering eXpert with Dynamic Data Selection
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.