A computational theory reduces agent evaluation errors and speeds up processes.
problem Efficient evaluation of mini agents at reduced cost.
method Developed a computational theory and a meta-learner to handle heterogeneous agents.
result Reduced evaluation errors by 24.1% to 99.0% across various scenarios.
Progress in machine learning is measured by careful evaluation on problems of outstanding common interest. However, the proliferation of benchmark suites and environments, adversarial attacks, and other complications has diluted the basic evaluation model by overwhelming researchers with choices. Deliberate or accident…
This paper addresses the problem of evaluating learning systems in safety critical domains such as autonomous driving, where failures can have catastrophic consequences. We focus on two problems: searching for scenarios when learned agents fail and assessing their probability of failure. The standard method for agent e…
AI agents improve forecast combination in empirical economics.
problem Hidden researcher degrees of freedom in AI-generated code.
method Adapted agent-loop architecture to empirical economics, added holdout evaluation.
result Independent agent searches find better forecast methods than benchmarks.
AI agents improve forecast combination but require transparency.
problem AI coding agents increase flexibility in empirical economics, leading to hidden degrees of freedom.
method Adapted open-source agent-loop architecture to empirical economics workflow, adding post-search holdout evaluation.
result Multiple agent runs outperform standard benchmarks in rolling evaluation but not all on post-search holdout.
The Neural Testbed evaluates joint predictions of neural agents, revealing their limitations.
problem Evaluating the quality of joint predictions generated by neural agents.
method Developed an open-source benchmark (The Neural Testbed) to assess agents' marginal and joint predictions.
result Popular Bayesian deep learning agents perform poorly on joint predictions, even with accurate marginal predictions.
Benchmark evaluates LLM trading agents by masking identifiers to prevent memory leaks.
problem Evaluate LLM trading agents without relying on market memory or noise.
method Data-side masking protocol, Barra-style performance attribution framework.
result LLM agents' returns are largely explained by market and style exposure, not stock selection.
This paper extends off-policy reinforcement learning to the multi-agent case in which a set of networked agents communicating with their neighbors according to a time-varying graph collaboratively evaluates and improves a target policy while following a distinct behavior policy. To this end, the paper develops a multi-…
InvestorBench benchmarks LLM agents in financial tasks.
problem Lack of a comprehensive benchmark for LLM-based financial agents.
method Developed a benchmark with diverse financial tasks and datasets.
result Evaluated LLM agents' performance across various financial products and market environments.
This work benchmarks MARL algorithms in cooperative tasks.
problem Lack of evaluation tasks and criteria for comparing MARL algorithms.
method Systematic evaluation of three MARL algorithm classes in diverse cooperative tasks.
result Insights into the effectiveness of different learning approaches.
The paper proposes a new method to evaluate LLM agent responses using ECDF clustering.
problem The standard evaluation of LLM agent responses via majority voting obscures response quality and distribution.
method The paper introduces a novel evaluation framework based on ECDF of cosine similarities and clustering of ECDFs using distances and k-medoids algorithm. result ECDF clustering reveals interpretable group structures in LLM responses, offering insights into agent settings.
Revealed preference theory studies the possibility of modeling an agent's revealed preferences and the construction of a consistent utility function. However, modeling agent's choices over preference orderings is not always practical and demands strong assumptions on human rationality and data-acquisition abilities. Th…
The paper shows how shared random seeds can reduce variance in machine learning evaluations.
problem The statistical structure of comparative evaluation under shared random seeds is not well understood.
method An extended learning-based multi-agent economic simulator was used to demonstrate the effects of shared random seeds on variance reduction.
result Pairing seeds can reduce variance in machine learning evaluations, especially when outcomes are positively correlated at the seed level.
IPO Finance Agent evaluates LLMs on SpaceX IPO due diligence, surpassing Finance Agent v2.
problem Evaluating language models on financial tasks like IPO due diligence.
method Introducing IPO Finance Agent and an evaluator-optimizer pipeline.
result The best-performing model reaches 79.4% accuracy at 0.30 USD per query.
New algorithm reduces multi-agent bandit regret by sharing data.
problem Designing efficient collaboration between multi-agent linear bandits.
method Bandit Adaptive Sample Sharing (BASS) algorithm, without assumptions on bandit parameters structure.
result Validated through theoretical analysis and empirical evaluations, BASS outperforms current state-of-the-art.
MaxMax Q-Learning improves coordination in multi-agent reinforcement learning by refining action selection.
problem Relative over-generalization in decentralized multi-agent reinforcement learning.
method MaxMax Q-Learning employs iterative sampling and evaluation of potential next states to refine approximations of ideal state transitions.
result MaxMax Q-Learning frequently outperforms existing baselines, demonstrating enhanced convergence and sample efficiency.
ISMCTS-BR learns best responses in large games, approximating worst-case performance.
problem Learning robustness to worst-case outcomes in large games.
method ISMCTS-BR, a scalable search-based algorithm for deep reinforcement learning.
result ISMCTS-BR approximates worst-case performance in large games.
Develops methods to create consistent surrogate models for agent-based simulators.
problem High computational costs and misjudgment of interventions in agent-based models.
method Causal abstractions to learn interventionally consistent surrogate models.
result Surrogates trained for interventional consistency closely mimic the agent-based model's behavior under interventions.
This paper presents a novel approach to the technical analysis of wireheading in intelligent agents. Inspired by the natural analogues of wireheading and their prevalent manifestations, we propose the modeling of such phenomenon in Reinforcement Learning (RL) agents as psychological disorders. In a preliminary step tow…
IPO Finance Agent extends Finance Agent v2 for SpaceX S-1 filings, improving accuracy and cost-efficiency.
problem Evaluating IPO due diligence tasks with long-form documents.
method Extended task domain, improved agentic harness with contextual retrieval, automated rubric generation.
result Best-performing model reaches 79.8% accuracy, cost-efficient model at 77.2% with 0.05 USD per query.
AgensFlow learns multi-agent coordination policies from experience.
problem Difficult coordination choices in multi-agent systems built on LLMs.
method Online policy learning from repeated trajectories, treating decisions as learnable.
result Learned routing improves coordination-heavy workflows over static wiring.
We consider the issue of multiple agents learning to communicate through reinforcement learning within partially observable environments, with a focus on information asymmetry in the second part of our work. We provide a review of the recent algorithms developed to improve the agents' policy by allowing the sharing of …
QMIX combines per-agent values to create decentralised policies.
problem Training decentralised policies from centralised learning.
method QMIX uses a mixing network to estimate joint action-values as a monotonic combination of per-agent values.
result QMIX significantly outperforms existing methods on the StarCraft Multi-Agent Challenge (SMAC).
AI-Trader benchmarks LLMs in live financial markets, revealing poor trading performance.
problem Challenges in real-time financial decision-making by autonomous agents.
method Fully automated, live evaluation benchmark with minimal human intervention.
result General intelligence does not translate to effective trading, highlighting limitations.
CausalGame benchmarks LLM agents' causal thinking in games.
problem Evaluating causal thinking in AI Scientists with LLMs.
method Interactive games with 14 scenarios incorporating selection bias, measurement error, and hidden confounders.
result None of the 30 LLM agents demonstrated reliable causal thinking, with the best model achieving only 68.0% survival.
Algorithm improves learning by integrating diverse agents' behaviors.
problem Lack of social learning in reinforcement learning algorithms.
method Free energy approach for social bandit learning.
result Algorithm converges to optimal policy and enhances learning.
Flexible decentralized MARL framework for cooperative multi-agent learning.
problem Complexity and impracticality of centralized MARL in complicated applications.
method Flexible fully-decentralized actor-critic MARL framework using primal-dual hybrid gradient descent.
result Competitive performance in large-scale cooperative multi-agent environments.
We show how a multi-agent simulator can support two important but distinct methods for assessing a trading strategy: Market Replay and Interactive Agent-Based Simulation (IABS). Our solution is important because each method offers strengths and weaknesses that expose or conceal flaws in the subject strategy. A key weak…
A new framework uses multi-agent reinforcement learning for evaluating policies in two-sided markets.
problem Evaluating the effects of different policies in two-sided markets with spatial and temporal interference.
method Introduces a multi-agent reinforcement learning (MARL) framework to address policy evaluation challenges in large-scale fleet management.
result Proposes novel estimators for mean outcomes under different products that are consistent despite high-dimensionality.
Memory is an important aspect of intelligence and plays a role in many deep reinforcement learning models. However, little progress has been made in understanding when specific memory systems help more than others and how well they generalize. The field also has yet to see a prevalent consistent and rigorous approach f…
This work develops a fully decentralized multi-agent algorithm for policy evaluation. The proposed scheme can be applied to two distinct scenarios. In the first scenario, a collection of agents have distinct datasets gathered following different behavior policies (none of which is required to explore the full state spa…
Foresight Arena benchmarks AI forecasting on real-world markets, isolating predictive edge.
problem Evaluating AI forecasting ability in real-world markets is challenging due to overfitting, centralized trust, and conflated metrics.
method Permissionless, on-chain benchmark using probabilistic forecasts, commit-reveal protocol, and smart contracts.
result Demonstrates the need for 350 predictions to reliably distinguish agents of different skill levels.
Agent Trading Arena trains LLMs in real-time financial markets to improve numerical reasoning.
problem Limited real-world training for LLMs in financial markets.
method Virtual zero-sum stock market with competitive multi-agent trading.
result LLMs perform better with chart-based visualizations and a reflection module.
FinSphere improves stock analysis quality with AI and expert-curated data.
problem Lack of objective evaluation metrics and depth in stock analysis by FinLLMs.
method Developed AnalyScore, curated Stocksis dataset, and FinSphere AI agent.
result FinSphere outperforms general and domain-specific LLMs in generating high-quality stock analysis reports.
Paper introduces metrics for evaluating multi-agent policies using best response dynamics.
problem Evaluation and ranking of multi-agent policies in reinforcement learning.
method Adopting strict best response dynamics (SBRD) to model selfish behaviors, proposing perturbed SBRD for dynamic and non-stationary settings.
result Proposed perturbed SBRD can observe policies with maximum metrics and differ from optimal by any given tolerance.
A test measures artificial agents' human-like behavior in video games.
problem Measuring the believability of artificial agents' human-like behavior.
method Developed a non-parametric two-sample hypothesis test.
result The p-value correlates with human judgment of human-like behavior. Evaluation of deep reinforcement learning (RL) is inherently challenging. In particular, learned policies are largely opaque, and hypotheses about the behavior of deep RL agents are difficult to test in black-box environments. Considerable effort has gone into addressing opacity, but almost no effort has been devoted t…
Method models other agents' behaviors without requiring direct observation.
problem Understanding and interacting effectively with other agents in reinforcement learning.
method Extracts representations from local observations of the controlled agent using encoder-decoder architectures.
result The method achieves higher returns than baseline methods in multi-agent environments.
The study addresses biases in evaluating molecular optimization methods and proposes methods to reduce these biases.
problem Biases in in silico evaluation of molecular optimization methods.
method Discussion and empirical investigation of bias reduction methods for predictor misspecification and sample reuse.
result Empirical investigation of bias reduction methods for predictor misspecification and sample reuse.
In the last few years, deep multi-agent reinforcement learning (RL) has become a highly active area of research. A particularly challenging class of problems in this area is partially observable, cooperative, multi-agent learning, in which teams of agents must learn to coordinate their behaviour while conditioning only…
RADIAL-RL improves deep RL agents' robustness against adversarial attacks.
problem Vulnerability of deep reinforcement learning agents to small adversarial perturbations.
method RADIAL-RL, a principled framework for training robust reinforcement learning agents.
result RADIAL-RL-trained agents consistently outperform prior methods in robustness tests.
Recent trends in Agent Computational Economics research, envelop a government agent in the model of the economy, whose decisions are based on learning algorithms. In this paper we try to evaluate the performance of simulated annealing in this context, by considering a model proposed earlier in the literature, which has…
Dynamic sentiment analysis improves stock trading strategies.
problem Improving stock trading performance through dynamic agent switching.
method Combining sentiment analysis with deep reinforcement learning and dynamic agent switching.
result Dynamic sentiment-based trading strategy outperforms traditional methods.
A trading system uses LLMs to adapt to volatile crypto markets.
problem Volatility and market sentiment in cryptocurrencies make traditional models ineffective.
method Specialized LLM agents for technical analysis, sentiment evaluation, and decision-making; verbal feedback for continuous improvement.
result Agents outperform buy-and-hold strategy with consistent gains across market phases.
Simple model-based reinforcement learning outperforms model-free methods in complex tasks.
problem Lagging performance of model-based reinforcement learning agents in non-trivial environments.
method Combining soft value estimates with stochastic value gradients.
result Simple model-based agents achieve state-of-the-art results in a high-dimensional humanoid control task.
Reinforcement learning (RL) algorithms allow agents to learn skills and strategies to perform complex tasks without detailed instructions or expensive labelled training examples. That is, RL agents can learn, as we learn. Given the importance of learning in our intelligence, RL has been thought to be one of key compone…
The study classifies and imitates trading agents in financial markets.
problem Classifying and imitating trading agents in continuous double auctions.
method Developed an agent-based model for trading, applied opponent modeling for classification, and used behavioral cloning for imitation.
result Techniques for classification and imitation were experimentally compared and evaluated.
MARS-Gym framework for marketplaces to train and evaluate recommender systems.
problem Challenges in designing, training, and evaluating recommender systems in marketplaces.
method Open-source framework for Reinforcement Learning agents in marketplaces.
result Empowers researchers and engineers to quickly build and evaluate agents for recommendations.