Study Examines Sources of Variability in AI Agent Outputs Across Multiple Runs
A preprint published on arXiv examines why agentic AI systems behave inconsistently across runs, even when given identical requests. The paper distinguishes between intrinsic variability—stemming from probabilistic token sampling during text generation—and extrinsic variability from factors like live data feeds, infrastructure differences, and numerical precision. Understanding these distinct sources matters for anyone seeking to audit, reproduce, or deploy AI agents reliably.
Researchers have published a preprint on arXiv analyzing the sources of behavioral variability in agentic AI systems—systems that plan, call tools, and update state in iterative loops. The paper argues that a key intrinsic source of variability is token sampling: when a model generates text, it converts scores over possible next tokens into probabilities and samples from them using a pseudo-random number generator, meaning small differences in sampled tokens can cascade into entirely different tool calls, code edits, or final answers. Beyond this, the authors identify extrinsic sources of variability including changing environments, live data, serving infrastructure, batch processing effects, and floating-point numerical details. A central contribution of the paper is separating these layers, which are often conflated in discussions of AI stochasticity. The authors clarify that even deterministic execution—fixing the random seed—does not guarantee identical behavior in real deployed settings due to these extrinsic factors. The work has implications for AI auditing, reproducibility standards, and the governance of autonomous AI systems.
What's missing
The paper is a preprint and has not yet undergone peer review, so its framework and claims have not been independently validated. The study appears primarily theoretical and taxonomic; it is unclear whether the authors empirically quantify the relative contribution of each variability source across real deployed systems.
What different sources said
- arXiv cs.AICenter
Impatient Users Confuse AI Agents: High-fidelity Simulations of Human Traits for Testing Agents
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.