arXiv subjects

Honglin Bao

Publications and source records attributed to Honglin Bao.

13 recordsLinked to original sources

arXiv2026

Contemporary AI lacks the imagination to diverge or negate in science

Bold claims that AI will accelerate scientific discovery have raced ahead of evidence from working scientists, yet large-scale, scientist-in-the-loop evidence is scarce. Here we mount the largest evaluation to date, inviting authors of 121,640 recent preprints in biology, medicine, chemistry, and social science to judge large language model (LLM)-generated ideas derived from their own papers. 6,749 representative scientists returned 25,139 rating sets on novelty, feasibility, probability of being true, and favorability of adoption. Three patterns emerge. First, non-reasoning LLMs collapse into a narrow "hivemind" of similar ideas while reasoning models explore a wider hypothesis space, but no model spontaneously proposes null hypotheses, a move humans make more freely. Second, scientists reward ideas resembling their own and prize probability over novelty, though social scientists tolerate risk more than life scientists; senior social scientists are the harshest critics, and their skepticism is earned, as LLMs falter most in pluralistic fields demanding context-aware interpretation and evolving theories. Third, automated evaluators, including LLM-as-a-judge and state-of-the-art (SOTA) models, agree weakly with expert judgment. Retrieval augmentation and scientist persona prompting yield marginal gains. A Qwen3-14B reward model we post-trained on human ratings captures nuances of taste, beats SOTA models by up to 27%, and closes the gap to the consistency of human peer reviewers. An analysis of 39 million papers from 2010 to 2025 links survey findings to macro-level patterns: following ChatGPT's release, null claims are sharply suppressed and ideas contract. Agent-based simulations further suggest that saturated fields should especially prize human uniqueness. For all the hype, today's AI for science remains a collaborator whose imagination and judgment benefit from human grounding.

cs.CY↗

arXiv2026

Agents' Last Exam

Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue that this gap is largely an evaluation problem: widely used benchmarks lack sustained performance measurement on real and economically valuable workflows. This paper introduces Agents' Last Exam (ALE), a benchmark designed to evaluate AI agents on long horizon, economically valuable, real world tasks with verifiable outcomes. Developed in collaboration with 250+ industry experts, ALE covers non-physical industries defined with reference to O*NET / SOC 2018 (the U.S. federal occupational taxonomy). It is organized around a task taxonomy with 55 sub fields grouped into 13 industry clusters covering 1K+ tasks. Current results show that the hardest tier remains far from saturated: across mainstream harness and backbone configurations, the average full pass rate is below 1%. ALE is designed as a living benchmark: its task pool grows continuously as new workflows and industries are onboarded. More broadly, ALE is intended not merely as another leaderboard, but as an instrument for closing the gap between benchmark success and GDP relevant impact.

cs.AI↗

arXiv2026

Narrative Flattening: How Post-Training Compresses Thematic, Affective, and Stylistic Variation in LLM Fiction

Large language models produce fluent fiction, yet their creative output is widely seen as flat. We ask where this quality originates in the training and whether it affects different domains of human fiction equally. We construct a matched story-continuation paradigm across StoryStar (public-platform), TMAS (prompt-guided), and The New Yorker (professional literary)-and compare continuations from four OLMo 32B checkpoints (Base, SFT, DPO, RLVR) against matched human text. Because these checkpoints share architecture, scale, tokenizer, and pretraining, the design isolates the post-training effect. We measure each continuation along three sentence-level dimensions: thematic motion, affective prevalence, and linguistic diversity. Across all three, post-training compresses dynamic variation: thematic transitions become more uniform, high-intensity emotions give way to neutrality, and stylistic diversity across stories shrinks. We term this progressive loss narrative flattening. The effect is directionally stable across story domains but gap size depends on the human baseline: professional literary fiction is compressed most, while public-platform and prompt-guided stories show smaller gaps, consistent with their human baselines sitting closer to the model's default rhythm. Post-trained endpoints converge across domains, suggesting alignment produces a continuation regime largely insensitive to the source domain's narrative texture.

cs.CL↗

arXiv2026

Building an Atlas of Social Experiments to Link Studies, Reconcile Conflicts, and Bridge Gaps

Social and behavioral science runs thousands of experiments each year, yet their findings rarely accumulate into a coherent map of what is known, what conflicts, and what remains missing. We introduce ExAtlas, a framework for turning an archive of experiments into an atlas: a structured map in which studies link, conflict, or leave bridgeable gaps. Given a target study, ExAtlas searches for prior studies that are locally close in treatment and outcome space and asks whether their observed effects can be composed to predict the target effect. This yields three cases. If the composition succeeds and agrees with the observed result, ExAtlas links the target to consistent prior evidence. If composition succeeds but disagrees, ExAtlas reconciles the conflict and proposes candidate moderators or higher-level theories that could explain it. If composition fails, ExAtlas proposes bridge experiments to close the gap. We provide an error bound for composition under local smoothness of the treatment-effect surface. On held-out targets certified as locally supported, ExAtlas recovers effect direction in 98.6% of cases. Human evaluations further suggest that its proposed bridge experiments are plausible and exhibit connectedness, and that its conflict explanations are useful for theory generation. These results suggest that the archive of social experiments contains more latent structure than current practice extracts -- and that making this structure explicit can guide both future theory and future experimentation.

cs.CY↗

arXiv2025

Mapping Overlaps in Benchmarks through Perplexity in the Wild

We introduce benchmark signatures to characterize the capacity demands of LLM benchmarks and their overlaps. Signatures are sets of salient tokens from in-the-wild corpora whose model token perplexity, reflecting training exposure, predicts benchmark performance. We extract them via stepwise forward selection with linear regression in a meta-evaluation spanning 32 LLMs and 89 benchmarks across diverse domains. We then analyze how these signatures relate to both the semantic similarity of benchmark questions and the correlation structure of model performance. While performance correlations are uniformly high and semantic overlaps stay in a narrow mid-range, benchmark signatures reveal more nuanced structure. For instance, they uncover substantial overlap between benchmarks in knowledge and reasoning tasks, whereas benchmarks in culture- and humanity-oriented domains show low similarity with each other. Unlike raw performance correlations, which are influenced by benchmark-orthogonal factors such as question formats, signatures are robust to such confounds. We further identify cross-functional overlaps between logic, math, language, instruction following, and cultural/world modeling, with coding emerging as the most isolated function, interacting only moderately with the ability of detecting missing information. Qualitative analysis shows that only the knowledge signature aligns with actual knowledge, suggesting that LLM semantic organization may differ from human conceptual structure. Together, these findings offer insights into benchmark validity, LLM sensitivities, and the landscape of interconnected LLM capacities. We have open-sourced the code and data in this https://github.com/siyangwu1/Benchmark-Signature-Repository.

cs.AI↗

arXiv2025

Persistence Paradox in Dynamic Science

Persistence is often regarded as a virtue in science. In this paper, however, we challenge this conventional view by highlighting its contextual nature, particularly how persistence can become a liability during periods of paradigm shift. We focus on the deep learning revolution catalyzed by AlexNet in 2012. Analyzing the 20-year career trajectories of over 5,000 scientists who were active in top machine learning venues during the preceding decade, we examine how their research focus and output evolved. We first uncover a dynamic period in which leading venues increasingly prioritized cutting-edge deep learning developments that displaced relatively traditional statistical learning methods. Scientists responded to these changes in markedly different ways. Those who were previously successful or affiliated with old teams adapted more slowly, experiencing what we term a rigidity penalty - a reluctance to embrace new directions leading to a decline in scientific impact, as measured by citation percentile rank. In contrast, scientists who pursued strategic adaptation - selectively pivoting toward emerging trends while preserving weak connections to prior expertise - reaped the greatest benefits. Taken together, our macro- and micro-level findings show that scientific breakthroughs act as mechanisms that reconfigure power structures within a field.

cs.DL↗

arXiv2025

Language Models Should be Used to Surface the Unwritten Code of Science and Society

This paper calls on the research community not only to investigate how human biases are inherited by large language models (LLMs) but also to explore how these biases in LLMs can be leveraged to make society's "unwritten code" - such as implicit stereotypes and heuristics - visible and accessible for critique. We introduce a conceptual framework through a case study in science: uncovering hidden rules in peer review - the factors that reviewers care about but rarely state explicitly due to normative scientific expectations. The idea of the framework is to push LLMs to speak out their heuristics through generating self-consistent hypotheses - why one paper appeared stronger in reviewer scoring - among paired papers submitted to 46 academic conferences, while iteratively searching deeper hypotheses from remaining pairs where existing hypotheses cannot explain. We observed that LLMs' normative priors about the internal characteristics of good science extracted from their self-talk, e.g., theoretical rigor, were systematically updated toward posteriors that emphasize storytelling about external connections, such as how the work is positioned and connected within and across literatures. Human reviewers tend to explicitly reward aspects that moderately align with LLMs' normative priors (correlation = 0.49) but avoid articulating contextualization and storytelling posteriors in their review comments (correlation = -0.14), despite giving implicit reward to them with positive scores. These patterns are robust across different models and out-of-sample judgments. We discuss the broad applicability of our proposed framework, leveraging LLMs as diagnostic tools to amplify and surface the tacit codes underlying human society, enabling public discussion of revealed values and more precisely targeted responsible AI.

cs.CY↗

arXiv2025

Missing vs. Unused Knowledge Hypothesis for Language Model Bottlenecks in Patent Understanding

While large language models (LLMs) excel at factual recall, the real challenge lies in knowledge application. A gap persists between their ability to answer complex questions and their effectiveness in performing tasks that require that knowledge. We investigate this gap using a patent classification problem that requires deep conceptual understanding to distinguish semantically similar but objectively different patents written in dense, strategic technical language. We find that LLMs often struggle with this distinction. To diagnose the source of these failures, we introduce a framework that decomposes model errors into two categories: missing knowledge and unused knowledge. Our method prompts models to generate clarifying questions and compares three settings -- raw performance, self-answered questions that activate internal knowledge, and externally provided answers that supply missing knowledge (if any). We show that most errors stem from failures to deploy existing knowledge rather than from true knowledge gaps. We also examine how models differ in constructing task-specific question-answer databases. Smaller models tend to generate simpler questions that they, and other models, can retrieve and use effectively, whereas larger models produce more complex questions that are less effective, suggesting complementary strengths across model scales. Together, our findings highlight that shifting evaluation from static fact recall to dynamic knowledge application offers a more informative view of model capabilities.

cs.CL↗

arXiv2025

Measuring Vogue in American Sociology (2011-2020)

This study investigates the social dynamics of knowledge production in American sociology. Departing from traditional approaches focused on citations, co-authorship, and faculty hiring, we introduce a method capturing the dynamics of networks inferred from text to explore which ideas gain traction (a.k.a vogue). Drawing on sociology doctoral dissertations and journal abstracts, we trace the movement of word pairs between peripheral and core semantic networks to uncover dominant themes and methodological trajectories. Our findings demonstrate that regional location and institutional prestige play critical roles in shaping the production and adoption of research trends across 114 sociology PhD-granting institutions in the United States. We show that applied research topics, such as crime and health, despite being perceived as less prestigious than theoretically oriented subjects, serve as the primary driving force behind the emergence and diffusion of trends within the discipline. This work sheds light on the institutional mechanisms that govern knowledge production, demonstrating that sociology's intellectual landscape is not dictated by simple top-down diffusion from elite institutions but is instead structured by the contextual and institutional factors that facilitate specialization and segmentation.

cs.SI↗

arXiv2024

From Division to Unity: A Large-Scale Study on the Emergence of Computational Social Science, 1990-2021

We present a comprehensive study on the emergence of Computational Social Science (CSS) - an interdisciplinary field leveraging computational methods to address social science questions - and its impact on adjacent social sciences. We trained a robust CSS classifier using papers from CSS-focused venues and applied it to 11 million papers spanning 1990 to 2021. Our analysis yielded three key findings. First, there were two critical inflections in the rise of CSS. The first occurred around 2005 when psychology, politics, and sociology began engaging with CSS. The second emerged in approximately 2014 when economics finally joined the trend. Sociology is currently the most engaged with CSS. Second, using the density of yearly knowledge embeddings constructed by advanced transformer models, we observed that CSS initially lacked a cohesive identity. From the early 2000s to 2014, however, it began to form a distinct cluster, creating boundaries between CSS and other social sciences, particularly in politics and sociology. After 2014, these boundaries faded, and CSS increasingly blended with the social sciences. Third, shared data-driven methods homogenized CSS papers across disciplines, with politics and economics showing the most alignment due to the combined influence of CSS and causal identification. Nevertheless, non-CSS papers in sociology, psychology, and politics became more divergent. Taken together, these findings highlight the dynamics of division and unity as new disciplines emerge within existing knowledge landscapes. A live demo of CSS evolution can be found in https://evolution-css.netlify.app/

cs.CY↗

arXiv2024

Where there's a will there's a way: ChatGPT is used more for science in countries where it is prohibited

Regulating AI is a key societal challenge, but which regulation methods are effective is unclear. This study measures the effectiveness of restricting AI services geographically, focusing on ChatGPT. OpenAI restricts ChatGPT access in several countries, including China and Russia. If restrictions are effective, ChatGPT use should be minimal in these countries. We measured use with a classifier based on distinctive word usage found in early versions of ChatGPT, e.g. "delve." We trained the classifier on pre- and post-ChatGPT "polished" abstracts and found it outperformed GPTZero and ZeroGPT on validation sets, including papers with self-reported AI use. Applying the classifier to preprints from Arxiv, BioRxiv, and MedRxiv showed ChatGPT was used in about 12.6% of preprints by August 2023, with 7.7% higher usage in restricted countries. The gap appeared before China's first major legal LLM became widely available. To test the possibility that, due to high demand, use in restricted countries would have been even higher without restrictions, we compared Asian countries with high expected demand (where English is not an official language) and found that use was higher in those with restrictions. ChatGPT use was correlated with higher views and downloads, but not citations or journal placement. Overall, restricting ChatGPT geographically has proven ineffective in science and possibly other domains, likely due to widespread workarounds.

cs.DL↗

arXiv2023

Do "bad" citations have "good" effects?

The scientific community discourages authors of research papers from citing papers that did not influence them. Such "rhetorical" citations are assumed to degrade the literature and incentives for good work. While a world where authors cite only substantively appears attractive, we argue that mandating substantive citing may have underappreciated consequences on the allocation of attention and dynamism in scientific literatures. We develop a novel agent-based model in which agents cite substantively and rhetorically. Agents first select papers to read based on their expected quality, read them and observe their actual quality, become influenced by those that are sufficiently good, and substantively cite them. Next, agents fill any remaining slots in the reference lists by (rhetorically) citing papers that support their narrative, regardless of whether they were actually influential. By turning rhetorical citing on-and-off, we find that rhetorical citing increases the correlation between quality and citations, increases citation churn, and reduces citation inequality. This occurs because rhetorical citing redistributes some citations from a stable set of elite-quality papers to a more dynamic set with high-to-moderate quality and high rhetorical value. Increasing the size of reference lists, often seen as an undesirable trend, amplifies the effects. In sum, rhetorical citing helps deconcentrate attention and makes it easier to displace incumbent ideas, so whether it is indeed undesirable depends on the metrics used to judge desirability.

cs.DL↗

arXiv2021

Evolution of Cooperative Hunting in Artificial Multi-layered Societies

The complexity of cooperative behavior is a crucial issue in multiagent-based social simulation. In this paper, an agent-based model is proposed to study the evolution of cooperative hunting behaviors in an artificial society. In this model, the standard hunting game of stag is modified into a new situation with social hierarchy and penalty. The agent society is divided into multiple layers with supervisors and subordinates. In each layer, the society is divided into multiple clusters. A supervisor controls all subordinates in a cluster locally. Subordinates interact with rivals through reinforcement learning, and report learning information to their corresponding supervisor. Supervisors process the reported information through repeated affiliation-based aggregation and by information exchange with other supervisors, then pass down the reprocessed information to subordinates as guidance. Subordinates, in turn, update learning information according to guidance, following the "win stay, lose shift" strategy. Experiments are carried out to test the evolution of cooperation in this closed-loop semi-supervised emergent system with different parameters. We also study the variations and phase transitions in this game setting.

cs.CY↗