Camilo Chacón Sartori

Camilo Chacón Sartori, José H. García, Andrei Voicu Tomut et al.

In the context of algorithms for problem solving, procedural knowledge -- the know-how of algorithm design and operator composition -- remains implicit in code, lost between runs, and must be re-engineered for each new domain. Knowledge graphs (KGs) have proven effective for organizing declarative knowledge, yet current KG paradigms provide limited support for representing procedural knowledge as executable, learnable graph structures. We introduce \textit{Generative Executable Algorithm Knowledge Graphs} (GEAKG), a class of KGs whose nodes store executable operators, whose edges encode learned composition patterns, and whose traversal generates solutions. A GEAKG is \emph{generative} (topology and operators are synthesized by a Large Language Model), \emph{executable} (every node is runnable code), and \emph{transferable} (learned patterns generalize zero-shot across domains). The framework is domain-agnostic at the engine level: the same three-layer architecture and Ant Colony Optimization (ACO)-based learning engine can be instantiated across domains, parameterized by a pluggable ontology (\texttt{RoleSchema}). Two case studies -- sharing no domain-specific framework code -- provide concrete evidence for this framework hypothesis: (1)~Neural Architecture Search across 70 cross-dataset transfer pairs on two tabular benchmarks, and (2)~Combinatorial Optimization, where knowledge learned on the Traveling Salesman Problem transfers zero-shot to scheduling and assignment domains. Taken together, the results support that algorithmic expertise can be explicitly represented, learned, and transferred as executable knowledge graphs.

Camilo Chacón Sartori, José H. García

Retrieval-augmented generation (RAG) systems are often compared by asking a large language model (LLM) judge which answer is better. For multi-hop RAG, this has become a measurement problem as much as a modeling problem: the same score can reflect retrieval quality, answer length, lexical overlap, or a statistical test that ignores clustered data. We ask what happens when these choices are made explicit. We propose a minimum measurement standard for LLM-as-a-judge comparisons in RAG. The standard fixes the top-100 candidate pool, evidence budget, answer cap, generator, and prompt; it also requires pre-registered hypotheses, cluster-aware inference, an exact cluster sign-flip check when feasible, and second-judge replication. Clustered benchmarks can overstate progress; the field should adopt this standard. We stress-test it with Genetic Algorithm Decoder for Multi-hop Evidence Composition (GADMEC), an evolutionary evidence selector, on 400 multi-hop questions in computer science/machine learning (CS/ML) and Materials Science. The protocol changes the empirical story. A binomial test makes all four semantic-baseline comparisons look significant; cluster-aware inference leaves only one Bonferroni-significant result. BM25 beats pure semantic GADMEC under the same budget, while a lexical-semantic hybrid recovers in CS/ML and narrows the Materials Science gap.

Camilo Chacón Sartori, Guillem Rodríguez Corominas

Can an LLM learn how an optimizer behaves -- and use that knowledge to control it? We extend Code World Models (CWMs), LLM-synthesized Python programs that predict environment dynamics, from deterministic games to stochastic combinatorial optimization. Given suboptimal trajectories of $(1{+}1)$-$\text{RLS}_k$, the LLM synthesizes a simulator of the optimizer's dynamics; greedy planning over this simulator then selects the mutation strength $k$ at each step. On \lo{} and \onemax{}, CWM-greedy performs within 6\% of the theoretically optimal policy -- without ever seeing optimal-policy trajectories. On \jump{$_k$}, where a deceptive valley causes all adaptive baselines to fail (0\% success rate), CWM-greedy achieves 100\% success rate -- without any collection policy using oracle knowledge of the gap parameter. On the NK-Landscape, where no closed-form model exists, CWM-greedy outperforms all baselines across fifteen independently generated instances ($36.94$ vs.\ $36.32$; $p<0.001$) when the prompt includes empirical transition statistics. The CWM also outperforms DQN in sample efficiency (200 offline trajectories vs.\ 500 online episodes), success rate (100\% vs.\ 58\%), and generalization ($k{=}3$: 78\% vs.\ 0\%). Robustness experiments confirm stable synthesis across 5 independent runs.

Camilo Chacón Sartori

When an agent can articulate why something works, we typically take this as evidence of genuine understanding. This presupposes that effective action and correct explanation covary, and that coherent explanation reliably signals both. I argue that this assumption fails for contemporary Large Language Models (LLMs). I introduce what I call the Bidirectional Coherence Paradox: competence and grounding not only dissociate but invert across epistemic conditions. In low-observability domains, LLMs often act successfully while misidentifying the mechanisms that produce their success. In high-observability domains, they frequently generate explanations that accurately track observable causal structure yet fail to translate those diagnoses into effective intervention. In both cases, explanatory coherence remains intact, obscuring the underlying dissociation. Drawing on experiments in compiler optimization and hyperparameter tuning, I develop the Epistemic Triangle, a model of how priors, signals, and domain knowledge interact under varying observability. The results suggest that neither behavioral success nor explanatory accuracy alone suffices for attributing understanding. I argue that evaluating artificial epistemic agents requires a tripartite framework -- coherence, grounding, and a proper basing relation linking explanation to action. The systematic separation of knowing-that and knowing-how in LLMs thus challenges assumptions inherited from both epistemology and current AI evaluation practice.

Camilo Chacón Sartori

When multiple LLM-based code agents independently implement parts of the same class, they must agree on shared internal representations, even when the specification leaves those choices implicit. We study this coordination problem across 51 class-generation tasks, progressively stripping specification detail from full docstrings (L0) to bare signatures (L3), and introducing opposing structural biases (lists vs. dictionaries) to stress-test integration. Three findings emerge. First, a persistent specification gap: two-agent integration accuracy drops from 58% to 25% as detail is removed, while a single-agent baseline degrades more gracefully (89% to 56%), leaving a 25--39 pp coordination gap that is consistent across two Claude models (Sonnet, Haiku) and three independent runs. Second, an AST-based conflict detector achieves 97% precision at the weakest specification level without additional LLM calls, yet a factorial recovery experiment shows that restoring the full specification alone recovers the single-agent ceiling (89%), while providing conflict reports adds no measurable benefit. Third, decomposing the gap into coordination cost (+16 pp) and information asymmetry (+11 pp) suggests that the two effects are independent and approximately additive. The gap is not merely a consequence of hidden information, but reflects the difficulty of producing compatible code without shared decisions. These results support a specification-first view of multi-agent code generation: richer specifications are both the primary coordination mechanism and the sufficient recovery instrument.

Camilo Chacón Sartori

Safety benchmarks often buy scalability by fixing the prompt, the language, and the turn structure. For emotional-support chatbots, that bargain hides precisely where safety failures emerge: across a multilingual, multi-turn crisis conversation. We present EMPATH, a benchmark for safety evaluation of emotional-support chatbots. An auditor model role-plays help-seeking users, generating multi-turn conversations from 140 seed instructions and 34 personas. A judge model scores each full transcript against 19 metrics across five dimensions: crisis handling, therapeutic quality, conversational integrity, emotional safety, and cultural adaptation. EMPATH is built for Mexican Spanish and US English; the studies reported here run in Mexican Spanish. Auditor and judge are drawn from different model families, and the judge is treated as an instrument to be calibrated rather than trusted. A strict per-criterion rubric reveals material score inflation on 10 of the 19 metrics and restores discrimination. We study the measurement properties of the benchmark through judge calibration and cross-family inter-judge agreement. We also illustrate EMPATH on three frontier models, one of them open-weight. Aggregate scores sit within 0.74 points of one another, but per-metric profiles diverge by up to six points in model-specific places. Under the standard rubric, both the ranking and the weak spots are stable across a second, cross-family judge: 93% of scores fall within plus or minus 1. A five-run test-retest adds a second axis: even the steadiest model swings from 2 to 10 on a crisis metric across identical re-runs, and deepseek-v4-pro returns a different conversation on every run even at temperature 0. Run-to-run reliability is therefore a per-model safety property, not noise to average away. EMPATH is system-agnostic; the pipeline, seeds, personas, and rubrics are released for reuse.

Camilo Chacón Sartori, Christian Blum, Gabriela Ochoa

The ability of Large Language Models (LLMs) to generate high-quality text and code has fuelled their rise in popularity. In this paper, we aim to demonstrate the potential of LLMs within the realm of optimization algorithms by integrating them into STNWeb. This is a web-based tool for the generation of Search Trajectory Networks (STNs), which are visualizations of optimization algorithm behavior. Although visualizations produced by STNWeb can be very informative for algorithm designers, they often require a certain level of prior knowledge to be interpreted. In an attempt to bridge this knowledge gap, we have incorporated LLMs, specifically GPT-4, into STNWeb to produce extensive written reports, complemented by automatically generated plots, thereby enhancing the user experience and reducing the barriers to the adoption of this tool by the research community. Moreover, our approach can be expanded to other tools from the optimization community, showcasing the versatility and potential of LLMs in this field.

Camilo Chacón Sartori, Christian Blum

The integration of Large Language Models (LLMs) into optimization has created a powerful synergy, opening exciting research opportunities. This paper investigates how LLMs can enhance existing optimization algorithms. Using their pre-trained knowledge, we demonstrate their ability to propose innovative heuristic variations and implementation strategies. To evaluate this, we applied a non-trivial optimization algorithm, Construct, Merge, Solve and Adapt (CMSA) -- a hybrid metaheuristic for combinatorial optimization problems that incorporates a heuristic in the solution construction phase. Our results show that an alternative heuristic proposed by GPT-4o outperforms the expert-designed heuristic of CMSA, with the performance gap widening on larger and denser graphs. Project URL: https://imp-opt-algo-llms.surge.sh/

Camilo Chacón Sartori, Christian Blum

Large Language Models (LLMs) have shown notable potential in code generation for optimization algorithms, unlocking exciting new opportunities. This paper examines how LLMs, rather than creating algorithms from scratch, can improve existing ones without the need for specialized expertise. To explore this potential, we selected 10 baseline optimization algorithms from various domains (metaheuristics, reinforcement learning, deterministic, and exact methods) to solve the classic Travelling Salesman Problem. The results show that our simple methodology often results in LLM-generated algorithm variants that improve over the baseline algorithms in terms of solution quality, reduction in computational time, and simplification of code complexity, all without requiring specialized optimization knowledge or advanced algorithmic implementation skills.

Camilo Chacón Sartori

With the rise of generative AI (GenAI), Large Language Models are increasingly employed for code generation, becoming active co-authors alongside human programmers. Focusing specifically on this application domain, this paper articulates distinct ``Architectures of Error'' to ground an epistemic distinction between human and machine code generation. Examined through their shared vulnerability to error, this distinction reveals fundamentally different causal origins: human-cognitive versus artificial-stochastic. To develop this framework and substantiate the distinction, the analysis draws critically upon Dennett's mechanistic functionalism and Rescher's methodological pragmatism. I argue that a systematic differentiation of these error profiles raises critical philosophical questions concerning semantic coherence, security robustness, epistemic limits, and control mechanisms in human-AI collaborative software development. The paper also utilizes Floridi's levels of abstraction to provide a nuanced understanding of how these error dimensions interact and may evolve with technological advancements. This analysis aims to offer philosophers a structured framework for understanding GenAI's unique epistemological challenges, shaped by these architectural foundations, while also providing software engineers a basis for more critically informed engagement.

Camilo Chacón Sartori, Christian Blum, Filippo Bistaffa

The fast advancement of Large Vision-Language Models (LVLMs) has shown immense potential. These models are increasingly capable of tackling abstract visual tasks. Geometric structures, particularly graphs with their inherent flexibility and complexity, serve as an excellent benchmark for evaluating these models' predictive capabilities. While human observers can readily identify subtle visual details and perform accurate analyses, our investigation reveals that state-of-the-art LVLMs exhibit consistent limitations in specific visual graph scenarios, especially when confronted with stylistic variations. In response to these challenges, we introduce VisGraphVar (Visual Graph Variability), a customizable benchmark generator able to produce graph images for seven distinct task categories (detection, classification, segmentation, pattern recognition, link prediction, reasoning, matching), designed to systematically evaluate the strengths and limitations of individual LVLMs. We use VisGraphVar to produce 990 graph images and evaluate six LVLMs, employing two distinct prompting strategies, namely zero-shot and chain-of-thought. The findings demonstrate that variations in visual attributes of images (e.g., node labeling and layout) and the deliberate inclusion of visual imperfections, such as overlapping nodes, significantly affect model performance. This research emphasizes the importance of a comprehensive evaluation across graph-related tasks, extending beyond reasoning alone. VisGraphVar offers valuable insights to guide the development of more reliable and robust systems capable of performing advanced visual graph analysis.