arXiv subjects

Claire E. Stevenson

Publications and source records attributed to Claire E. Stevenson.

12 recordsLinked to original sources

arXiv2026

Measuring Combinatorial Creativity in Large Language Models and Humans

Combinatorial creativity, the process of combining familiar ideas in unfamiliar ways, is one of the foremost engines of scientific, technological, and artistic progress. This work introduces Kombine, an LLM benchmark for combinatorial creativity tasking models to discover surprising, original, and useful associations, analogies, and blends between unrelated entities, and to use those analogies and blends to invent novel concepts by re-combining the parts of existing ones. In a large-scale study across 35 LLMs, we find that: (1) LLMs are 10-16 percentage points better at discovering associations and analogies than blends, and (2) exhibit an abstraction bottleneck on blending, in which even frontier LLMs struggle to find a shared abstraction between inputs and instantiate a coherent blended space. (3) LLMs also exhibit inventive homogeneity, in which one in five concepts invented for a given item reuses properties invented by another model. Moreover, (4) models are $14\times$ more likely to reuse other invented properties on blends than analogies, and (5) models from the same provider are $2.3\times$ more likely to reuse properties than those from different providers. Next, in a comparative study involving 75 humans on the same tasks, we find that (6) humans score below nearly every LLM on each task and (7) both humans and LLMs find blending most difficult, but association is easier than analogy for LLMs, while analogy is easier than association for humans. Lastly, (8) humans and LLMs access distinct regions of semantic space on each task, suggesting that combinatorial creativity could benefit from systematic human-AI collaboration to enhance conceptual diversity. In summary, we provide a framework to assess progress towards LLMs, AI systems, and human-AI collaborations more broadly, that can elicit combinatorial creativity in a setting relevant to scientific and technological innovation.

cs.AI↗

arXiv2026

Which Attention Heads are like the Human Head? Not the Ones that Compute

Brain-AI alignment is often interpreted as a sign that model and brain perform similar computations. Whether the aligned units are causally involved in model computation is rarely checked. On an abstract pattern-completion task (AAABAAA $\rightarrow$ B), we compare LLM attention-head representations with human EEG and test how ablating those heads affects task performance. Alignment and causation dissociate: brain-aligned heads contribute to performance, but their removal is substantially less disruptive than removal of heads selected via attribution patching. We compare two head sets that prior interpretability work defines without reference to the brain: concept vectors (CVs), which represent abstract patterns across formats, and function vectors (FVs), selected for their contribution to correct-answer prediction. Brain alignment shows little association with FV scores, while its association with CV scores varies across models. Among brain-aligned heads, we find recurring attention profiles: one emphasizes distinctive elements (novelty heads), the other repeating elements (repetition heads). The novelty family tracks salience and attends to the same elements that humans look at, yet its removal is less damaging than random ablation on average. Repetition heads contribute modestly to performance and are associated with abstract-pattern representation (CVs). Across 17 models spanning 3B-72B parameters, FV-ranked removal is substantially more disruptive than brain-ranked removal. Brain alignment thus captures how the model reads the stimulus, and only faintly captures how it represents the pattern and solves the task.

cs.AI↗

arXiv2026

Transformer See, Transformer Do: Copying as an Intermediate Step in Learning Analogical Reasoning

Analogical reasoning is a hallmark of human intelligence, enabling us to solve new problems by transferring knowledge from one situation to another. Yet, developing artificial intelligence systems capable of robust human-like analogical reasoning has proven difficult. In this work, we train transformers using Meta-Learning for Compositionality (MLC) on an analogical reasoning task (letter-string analogies) and assess their generalization capabilities. We find that letter-string analogies become learnable when guiding the models to attend to the most informative problem elements, induced by including copy tasks in the training data. Furthermore, generalization to new alphabets improves when models are trained with more heterogeneous datasets. For the best training run, our 3-layer encoder-decoder model performs on par with frontier models on our letter-string analogy datasets. The MLC approach also enables some generalization to compositions of trained transformations, but not to completely novel transformations. To understand how the model solves the analogies, we identify an algorithm for one analogy task that approximates the model's computations. We verify this using interpretability analyses and show that head-level attention weights causally affect model output, allowing steering between different analogy tasks. We discuss the implications of our findings for generalization capabilities of larger models and the parallels to human analogical reasoning.

cs.LG↗

arXiv2026

AGC-Bench: Measuring Artificial General Creativity

Creativity research has debated whether creativity is domain-specific (e.g., visual, writing, science), and if it is psychometrically separable from general intelligence. Both questions now apply to LLMs, but a unified benchmark of AI creativity remains elusive. We introduce AGC-Bench, an artificial general creativity benchmark built from a systematic review of the AI creativity literature (3,101 papers screened, 497 benchmarks identified), paired with an agentic harness that converts idiosyncratic codebases into HELM-standardized benchmarks. The first release covers 78 datasets spanning brainstorming, problem solving, STEM, narrative, figurative language, and humor. To address bias in LLM-as-judge, we apply Judge Response Theory -- a psychometric calibration of judge leniency/severity; we then fine-tune Qwen3-30B on the bias-corrected ratings of three frontier LLMs to produce AGC-Judge, an open-weight model that robustly scores new creativity benchmarks it was not trained on. Results reveal frontier models at the top of the AGC-Bench leaderboard, with open models close behind. LLMs show different creative strengths, ranking higher on some domains (e.g., writing) than others (e.g., scientific ideation). Extensive experiments yield three main findings. First, applying factor analysis across 83 LLMs, we recover a single creativity factor 'c', analogous to the 'g' factor of general intelligence, that explains 81.5% of variance, related to but separable from general knowledge/reasoning. Second, we show that prompting models to "be creative" boosts their performance far more than enabling reasoning, evidence that the benchmark tracks creativity over general ability. Third, on a human-matched subset, we find the top human still leads the top LLM on creativity. We release AGC-Bench with a public leaderboard, AGC-Judge, and human data as open infrastructure for measuring AI creativity at scale.

cs.CL↗

arXiv2026

Causality $\neq$ Invariance: Function and Concept Vectors in LLMs

Do large language models (LLMs) represent concepts abstractly, i.e., independent of input format? We revisit Function Vectors (FVs), compact representations of in-context learning (ICL) tasks that causally drive task performance. Across multiple LLMs, we show that FVs are not fully invariant: FVs are nearly orthogonal when extracted from different input formats (e.g., open-ended vs. multiple-choice), even if both target the same concept. We identify Concept Vectors (CVs), which carry more stable concept representations. Like FVs, CVs are composed of attention head outputs; however, unlike FVs, the constituent heads are selected using Representational Similarity Analysis (RSA) based on whether they encode concepts consistently across input formats. While these heads emerge in similar layers to FV-related heads, the two sets are largely distinct, suggesting different underlying mechanisms. Steering experiments reveal that FVs excel in-distribution, when extraction and application formats match (e.g., both open-ended in English), while CVs generalize better out-of-distribution across both question types (open-ended vs. multiple-choice) and languages. Our results show that LLMs do contain abstract concept representations, but these differ from those that drive ICL performance.

cs.CL↗

arXiv2025

Can Large Language Models generalize analogy solving like children can?

In people, the ability to solve analogies such as "body : feet :: table : ?" emerges in childhood, and appears to transfer easily to other domains, such as the visual domain "( : ) :: < : ?". Recent research shows that large language models (LLMs) can solve various forms of analogies. However, can LLMs generalize analogy solving to new domains like people can? To investigate this, we had children, adults, and LLMs solve a series of letter-string analogies (e.g., a b : a c :: j k : ?) in the Latin alphabet, in a near transfer domain (Greek alphabet), and a far transfer domain (list of symbols). Children and adults easily generalized their knowledge to unfamiliar domains, whereas LLMs did not. This key difference between human and AI performance is evidence that these LLMs still struggle with robust human-like analogical transfer.

cs.AI↗

arXiv2025

Large Language Models Show Signs of Alignment with Human Neurocognition During Abstract Reasoning

This study investigates whether large language models (LLMs) mirror human neurocognition during abstract reasoning. We compared the performance and neural representations of human participants with those of eight open-source LLMs on an abstract-pattern-completion task. We leveraged pattern type differences in task performance and in fixation-related potentials (FRPs) as recorded by electroencephalography (EEG) during the task. Our findings indicate that only the largest tested LLMs (~70 billion parameters) achieve human-comparable accuracy, with Qwen-2.5-72B and DeepSeek-R1-70B also showing similarities with the human pattern-specific difficulty profile. Critically, every LLM tested forms representations that distinctly cluster the abstract pattern categories within their intermediate layers, although the strength of this clustering scales with their performance on the task. Moderate positive correlations were observed between the representational geometries of task-optimal LLM layers and human frontal FRPs. These results consistently diverged from comparisons with other EEG measures (response-locked ERPs and resting EEG), suggesting a potential shared representational space for abstract patterns. This indicates that LLMs might mirror human brain mechanisms in abstract reasoning, offering preliminary evidence of shared principles between biological and artificial intelligence.

q-bio.NC↗

arXiv2025

Analogical Reasoning Inside Large Language Models: Concept Vectors and the Limits of Abstraction

Analogical reasoning relies on conceptual abstractions, but it is unclear whether Large Language Models (LLMs) harbor such internal representations. We explore distilled representations from LLM activations and find that function vectors (FVs; Todd et al., 2024) - compact representations for in-context learning (ICL) tasks - are not invariant to simple input changes (e.g., open-ended vs. multiple-choice), suggesting they capture more than pure concepts. Using representational similarity analysis (RSA), we localize a small set of attention heads that encode invariant concept vectors (CVs) for verbal concepts like "antonym". These CVs function as feature detectors that operate independently of the final output - meaning that a model may form a correct internal representation yet still produce an incorrect output. Furthermore, CVs can be used to causally guide model behaviour. However, for more abstract concepts like "previous" and "next", we do not observe invariant linear representations, a finding we link to generalizability issues LLMs display within these domains.

cs.CL↗

arXiv2025

Pencils to Pixels: A Systematic Study of Creative Drawings across Children, Adults and AI

Can we derive computational metrics to quantify visual creativity in drawings across intelligent agents, while accounting for inherent differences in technical skill and style? To answer this, we curate a novel dataset consisting of 1338 drawings by children, adults and AI on a creative drawing task. We characterize two aspects of the drawings -- (1) style and (2) content. For style, we define measures of ink density, ink distribution and number of elements. For content, we use expert-annotated categories to study conceptual diversity, and image and text embeddings to compute distance measures. We compare the style, content and creativity of children, adults and AI drawings and build simple models to predict expert and automated creativity scores. We find significant differences in style and content in the groups -- children's drawings had more components, AI drawings had greater ink density, and adult drawings revealed maximum conceptual diversity. Notably, we highlight a misalignment between creativity judgments obtained through expert and automated ratings and discuss its implications. Through these efforts, our work provides, to the best of our knowledge, the first framework for studying human and artificial creativity beyond the textual modality, and attempts to arrive at the domain-agnostic principles underlying creativity. Our data and scripts are available on GitHub.

cs.HC↗

arXiv2024

Do Large Language Models Solve ARC Visual Analogies Like People Do?

The Abstraction Reasoning Corpus (ARC) is a visual analogical reasoning test designed for humans and machines (Chollet, 2019). We compared human and large language model (LLM) performance on a new child-friendly set of ARC items. Results show that both children and adults outperform most LLMs on these tasks. Error analysis revealed a similar "fallback" solution strategy in LLMs and young children, where part of the analogy is simply copied. In addition, we found two other error types, one based on seemingly grasping key concepts (e.g., Inside-Outside) and the other based on simple combinations of analogy input matrices. On the whole, "concept" errors were more common in humans, and "matrix" errors were more common in LLMs. This study sheds new light on LLM reasoning ability and the extent to which we can use error analyses and comparisons with human development to understand how LLMs solve visual analogies.

cs.CL↗

arXiv2023

Solving ARC visual analogies with neural embeddings and vector arithmetic: A generalized method

Analogical reasoning derives information from known relations and generalizes this information to similar yet unfamiliar situations. One of the first generalized ways in which deep learning models were able to solve verbal analogies was through vector arithmetic of word embeddings, essentially relating words that were mapped to a vector space (e.g., king - man + woman = __?). In comparison, most attempts to solve visual analogies are still predominantly task-specific and less generalizable. This project focuses on visual analogical reasoning and applies the initial generalized mechanism used to solve verbal analogies to the visual realm. Taking the Abstraction and Reasoning Corpus (ARC) as an example to investigate visual analogy solving, we use a variational autoencoder (VAE) to transform ARC items into low-dimensional latent vectors, analogous to the word embeddings used in the verbal approaches. Through simple vector arithmetic, underlying rules of ARC items are discovered and used to solve them. Results indicate that the approach works well on simple items with fewer dimensions (i.e., few colors used, uniform shapes), similar input-to-output examples, and high reconstruction accuracy on the VAE. Predictions on more complex items showed stronger deviations from expected outputs, although, predictions still often approximated parts of the item's rule set. Error patterns indicated that the model works as intended. On the official ARC paradigm, the model achieved a score of 2% (cf. current world record is 21%) and on ConceptARC it scored 8.8%. Although the methodology proposed involves basic dimensionality reduction techniques and standard vector arithmetic, this approach demonstrates promising outcomes on ARC and can easily be generalized to other abstract visual reasoning tasks.

cs.AI↗

arXiv2023

Do large language models solve verbal analogies like children do?

Analogy-making lies at the heart of human cognition. Adults solve analogies such as \textit{Horse belongs to stable like chicken belongs to ...?} by mapping relations (\textit{kept in}) and answering \textit{chicken coop}. In contrast, children often use association, e.g., answering \textit{egg}. This paper investigates whether large language models (LLMs) solve verbal analogies in A:B::C:? form using associations, similar to what children do. We use verbal analogies extracted from an online adaptive learning environment, where 14,002 7-12 year-olds from the Netherlands solved 622 analogies in Dutch. The six tested Dutch monolingual and multilingual LLMs performed around the same level as children, with MGPT performing worst, around the 7-year-old level, and XLM-V and GPT-3 the best, slightly above the 11-year-old level. However, when we control for associative processes this picture changes and each model's performance level drops 1-2 years. Further experiments demonstrate that associative processes often underlie correctly solved analogies. We conclude that the LLMs we tested indeed tend to solve verbal analogies by association with C like children do.

cs.CL↗