Explore or Confirm? LLM Agents Misallocate Measurements Under Noise

Ranked by how close their title and abstract are to this paper’s. Select some to read them together.

Closest papers

1. Close Not relevant

   Efficient Exploration for LLMs

   Vikranth Dwaracherla, Seyed Mohammad Asghari, Botao Hao et al. · 2024 · Machine Learning · AI summary available

   Abstract

   We present evidence of substantial benefit from efficient exploration in gathering human feedback to improve large language models. In our experiments, an agent sequentially generates queries while fitting a reward model to the feedback received. Our best-performing agent generates queries using double Thompson sampling, with uncertainty represented by an epistemic neural network. Our results demonstrate that efficient exploration enables high levels of performance with far fewer queries. Further, both uncertainty estimation and the choice of exploration scheme play critical roles.

2. Close Not relevant

   Compressing the Validation Bottleneck: An Agentic Self-Driving Lab for Scientific Discovery

   Kyunghoon Hur, Chi-Hun Lee · 2026 · arXiv.org · Not summarized yet

   Abstract

   Agentic AI-for-Science can automate ideation, planning, and analysis, but final validation still depends on real experiments. A self-driving lab (SDL) can execute those experiments, yet the loop still has bottlenecks: the agent may spend too many rounds on low-value experiments, or each round may require a high-cost experiment. We target these two physical bottlenecks with one agent. First, a prior-aware agentic DOE loop uses domain knowledge and past results to propose feasible and informative next experiments, reducing trials-to-target. Second, a cost-aware surrogate agent predicts high-cost, high-resolution measurements from low-cost, low-resolution measurements. It chooses between a high…

3. Close Not relevant

   Invisibilization as a Source of Probability Bias in Language Models

   Christian Arnold, Ernesto Calvo, Guadalupe Gonzalez et al. · 2026 · Computation and Language · AI summary available

   In short

   Leading language models often refuse to assess reports based on the reporter’s description, producing racial gaps in information use that average-trust audits can miss.

4. Close Not relevant

   AlphaResearch: Accelerating New Algorithm Discovery with Language Models

   Zhaojian Yu, Kaiyue Feng, Yilun Zhao et al. · 2025 · arXiv.org · Not summarized yet

   Abstract

   LLMs have made significant progress in complex but easy-to-verify problems, yet they still struggle with discovering the unknown. In this paper, we present \textbf{AlphaResearch}, an autonomous research agent designed to discover new algorithms on open-ended problems by iteratively running the following steps: (1) propose new ideas (2) program to verify (3) optimize the research proposals. To synergize the feasibility and innovation of the discovery process, we construct a novel dual environment by combining the execution-based verifiable reward and reward from simulated real-world peer review environment in AlphaResearch. We construct \textbf{\dataset}, a set of questions that includes an e…

5. Close Not relevant

   AgentFly: Fine-tuning LLM Agents without Fine-tuning LLMs

   Huichi Zhou, Yihang Chen, Siyuan Guo et al. · 2025 · Machine Learning · AI summary available

   Abstract

   In this paper, we introduce a novel learning paradigm for adaptive Large Language Model (LLM) agents that eliminates the need for fine-tuning the underlying LLMs. Existing approaches are often either rigid, relying on static, handcrafted reflection workflows, or computationally intensive, requiring gradient updates of LLM model parameters. In contrast, our method enables low-cost continual adaptation via memory-based online reinforcement learning. We formalise this as a Memory-augmented Markov Decision Process (M-MDP), equipped with a neural case-selection policy to guide action decisions. Past experiences are stored in an episodic memory, either differentiable or non-parametric. The policy …

6. Close Cited by this paper Not relevant

   Can AI Scientist Agents Learn from Lab-in-the-Loop Feedback? Evidence from Iterative Perturbation Discovery

   G. Wainrib, B. Bodinier, Haitem Dakhli et al. · 2026 · arXiv.org · Not summarized yet

   Abstract

   Recent work has questioned whether large language models (LLMs) can perform genuine in-context learning (ICL) for scientific experimental design, with prior studies suggesting that LLM-based agents exhibit no sensitivity to experimental feedback. We shed new light on this question by carrying out 800 independently replicated experiments on iterative perturbation discovery in Cell Painting high-content screening. We compare an LLM agent that iteratively updates its hypotheses using experimental feedback to a zero-shot baseline that relies solely on pretraining knowledge retrieval. Access to feedback yields a $+53.4\%$ increase in discoveries per feature on average ($p = 0.003$). To test wheth…

7. Close Not relevant

   More Agents Is All You Need

   Junyou Li, Qin Zhang, Yangbin Yu et al. · 2024 · Computation and Language · AI summary available

   Abstract

   We find that, simply via a sampling-and-voting method, the performance of large language models (LLMs) scales with the number of agents instantiated. Also, this method is orthogonal to existing complicated methods to further enhance LLMs, while the degree of enhancement is correlated to the task difficulty. We conduct comprehensive experiments on a wide range of LLM benchmarks to verify the presence of our finding, and to study the properties that can facilitate its occurrence. Our code is publicly available at: \url{https://anonymous.4open.science/r/more_agent_is_all_you_need}.

8. Close Not relevant

   Fisher-R1: Training LLM Agents for Reliable Hypothesis Testing

   Jia-Cheng Miao, Jin Mu, Guan-Hua Chen et al. · 2026 · Not summarized yet

   Abstract

   Reliable hypothesis testing is the foundation of many empirical scientific claims. Large language model (LLM) agents are increasingly used to automate this process, as they can inspect datasets, generate code, and produce analyses end-to-end. However, we show that they frequently make subtle inferential errors that lead to incorrect conclusions despite correctly executed analyses. Existing benchmarks fail to capture this failure mode, as they rarely assess whether a reported p-value is statistically valid given the assumptions underlying the data. We address this gap by building P-Bench, a benchmark comprising 425 open-ended, realistic hypothesis-testing tasks spanning economics, biology, an…

Prior work it builds on

1. Close Not relevant

   AllocBench: Measuring Online Tool Allocation Capability in LLM Agents

   Daniel Wang, Andrew Xu · 2026 · arXiv.org · Not summarized yet

   Abstract

   Creating a reusable tool is an investment: an agent pays a fixed cost now in exchange for the potential of future reuse. Therefore, a user should prefer an agent that creates a small number of highly reusable tools, rather than many one-offs. We introduce a paired benchmark that tests whether LLM agents exhibit conscious allocation behavior under a fixed budget in two contexts: an abstract text-based formulation and a code-construction task. We find that every frontier model we test---Claude Haiku, Claude Opus, GPT-5.4-mini, and GPT-5.6 Sol---acts near-optimally in the abstract framing but fails to transfer this ability to script-writing. Through further experiments, we identify the particul…

2. Close Not relevant

   LLMs for Bayesian Optimization in Scientific Domains: Are We There Yet?

   Rushil Gupta, Jason S. Hartford, Bang Liu · 2025 · Conference on Empirical Methods in Natural Language Processing · Not summarized yet

   Abstract

   Large language models (LLMs) have recently been proposed as general-purpose agents for experimental design, with claims that they can perform in-context experimental design. We evaluate this hypothesis using both open- and closed-source instruction-tuned LLMs applied to genetic perturbation and molecular property discovery tasks. We find that LLM-based agents show no sensitivity to experimental feedback: replacing true outcomes with randomly permuted labels has no impact on performance. Across benchmarks, classical methods such as linear bandits and Gaussian process optimization consistently outperform LLM agents. We further propose a simple hybrid method, LLM-guided Nearest Neighbour (LLMNN…

3. Close Not relevant

   AgentHPOBench: A Benchmark For Evaluating LLM Agents as Sequential Hyperparameter Optimizers

   Tianyu Huai, Ting-Shuo Fan, Xin-Chi Chen et al. · 2026 · arXiv.org · Not summarized yet

   Abstract

   As LLMs evolve from code completion systems into autonomous scientific agents, evaluating their ability to conduct experiments has become increasingly important. Existing benchmarks typically focus on static code generation, paper replication, or final answer correctness, but do not directly assess whether agents can interpret experimental evidence and use it to guide subsequent hyperparameter decisions. To address this gap, we introduce AgentHPOBench, a sequential benchmark comprising 30 executable machine learning tasks across seven research categories. Each task begins with a validated baseline run, after which an agent performs several sequential interventions. At each step, the agent ob…

4. Close Not relevant

   Benchmarking and behavioral characterization of LLM agents for protein design

   Jeonghyeon Kim, P. Romero · 2026 · bioRxiv · Not summarized yet

   Abstract

   Large language models (LLMs) are increasingly deployed as agents for scientific discovery, but standardized frameworks for evaluating their performance and behavior in scientific workflows are lacking. Protein design provides a demanding test case because modern workflows combine stochastic generative models, structure prediction systems, and physics-based evaluation tools that require extensive candidate exploration and filtering. Here we introduce BioDesignBench, a benchmark of 76 expert-curated protein design tasks spanning antibodies, enzymes, fluorescent proteins, binders, and scaffolds, together with human and non-LLM baselines and behavioral metrics derived from tool-use traces. We ev…

5. Related Not relevant

   BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments

   Yusuf H. Roohani, Andrew H. Lee, Qian Huang et al. · 2024 · International Conference on Learning Representations · Not summarized yet

   Abstract

   Agents based on large language models have shown great potential in accelerating scientific discovery by leveraging their rich background knowledge and reasoning capabilities. In this paper, we introduce BioDiscoveryAgent, an agent that designs new experiments, reasons about their outcomes, and efficiently navigates the hypothesis space to reach desired solutions. We demonstrate our agent on the problem of designing genetic perturbation experiments, where the aim is to find a small subset out of many possible genes that, when perturbed, result in a specific phenotype (e.g., cell growth). Utilizing its biological knowledge, BioDiscoveryAgent can uniquely design new experiments without the nee…