Agentic AI systems are increasingly embedded in scientific workflows, including literature synthesis, hypothesis generation, tool use, writing, and peer review. However, existing surveys and evaluations often emphasize task performance while underexamining how humans and agents coordinate under different epistemic conditions of scientific work. This survey introduces **CSCW**×**Kuhn**, a coordination-aware framework that connects computer-supported cooperative work dimensions with distinct scientific contexts. Using this framework, we synthesize recent human–agent systems and benchmarks to examine how coordination demands and failure risks may vary across contexts. Our synthesis identifies five recurring failure modes: procedural drift, grounding substitution, evaluative flattening, proxy optimization, and epistemic black-boxing, and argues for evaluations that move beyond task accuracy and toward grounding, provenance, disagreement preservation, and long-horizon scientific reliability. Our corpus and coding artifact are available at https://github.com/jenlia024/CSCWxKuhn-artifact.

The rapid proliferation of Generative AI has created a pressing need for rigorous documentation standards that support transparency and governance. Manual creation of Model and Data Cards, however, is difficult to scale, and automated approaches lack the large-scale, high-fidelity benchmarks required for systematic evaluation. We introduce MetaGAI, a benchmark of 2,541 semantically matched document triplets constructed from academic papers, GitHub repositories, and Hugging Face artifacts. Reference cards are produced by a multi-agent pipeline of specialized Retriever, Generator, and Editor agents, and the pipeline is assessed through a four-part validation protocol. We further propose an evaluation framework that combines reference-based automated metrics with a multi-model LLM-as-a-Judge protocol, together with explicit agreement and sensitivity analyses. Among the evaluated baselines, Qwen3-30B-A3B-Instruct achieves the best observed cost–quality trade-off, and the results reveal a recurring tension between faithfulness and completeness. MetaGAI provides a testbed for benchmarking, training, and analyzing automated Model and Data Card generation at scale. Our data and code are available at https://github.com/haoxuan-unt2024/MetaGAI-Benchmark.

A core challenge for faithful LLM role-playing is sustaining consistent characterization throughout long, open-ended dialogues, as models frequently fail to recall and accurately apply their designated persona knowledge without explicit cues. To tackle this, we propose the Memory-Driven Role-Playing paradigm. Inspired by Stanislavski’s "emotional memory” acting theory, this paradigm frames persona knowledge as the LLM’s internal memory store, requiring retrieval and application based solely on dialogue context, thereby providing a rigorous test of depth and autonomous use of knowledge. Centered on this paradigm, we contribute: (1) MREval, a fine-grained evaluation framework assessing four memory-driven abilities—Anchoring, Selecting, Bounding, and Enacting; (2) MRPrompt, a prompting architecture that guides structured memory retrieval and response generation; and (3) MRBench, a bilingual (Chinese/English) benchmark for fine-grained diagnosis. The novel paradigm provides a comprehensive diagnostic for four-stage role-playing abilities across 12 LLMs. Crucially, experiments show that MRPrompt allows small models (e.g., Qwen3-8B) to match the performance of much larger closed-source LLMs (e.g., Qwen3-Max and GLM-4.7), and confirm that upstream memory gains directly enhance downstream response quality, validating the staged theoretical foundation.

Large Language Models (LLMs) achieve strong reasoning performance, but their robustness to realistic lexical corruption remains poorly understood. We evaluate four open-weight instruction-tuned models and frontier models across four reasoning benchmarks under keyboard noise, character swaps, and filler insertion. Character-level perturbations substantially degrade accuracy, especially on multi-step reasoning tasks, while filler insertion has little effect. We trace this asymmetry to Attention Diversion: lexical corruption fragments subword tokenization, and the resulting fragments attract disproportionate attention mass, concentrated in middle and final transformer layers. Length-matched controls confirm that fragmentation, not prompt length, drives the loss. A factorial intervention then shows why the damage is hard to undo: fragmentation corrupts token content and attention allocation together, and the two are coupled. Restoring clean attention while the content remains corrupted is actively harmful, restoring content alone is insufficient, and only restoring both recovers a substantial share of the gap. This coupling explains why inference-time strategies, including chain-of-thought prompting, spell-checking, self-repair, and stronger repair models, fail to consistently recover performance: each addresses one channel at a time. Code and data are available at https://github.com/Jiaqian-Janelle/Attention-Diversion.

Autonomous LLM agents increasingly operate in long-horizon, interactive settings where success depends on reusing experience accumulated over extended histories. However, existing agent memory systems are fundamentally constrained by text-context budgets: storing or revisiting raw trajectories is prohibitively token-expensive, while summarization and text-only retrieval trade token savings for information loss and fragmented evidence. To address this limitation, we propose Optical Context Retrieval Memory (OCR-Memory), a memory framework that leverages the visual modality as a high-density representation of agent experience, enabling retention of arbitrarily long histories with minimal prompt overhead at retrieval time. Specifically, OCR-Memory renders historical trajectories into images annotated with unique visual identifiers. OCR-Memory retrieves stored experience via a locate-and-transcribe paradigm that selects relevant regions through visual anchors and retrieves the corresponding verbatim text, avoiding free-form generation and reducing hallucination. Experiments on long-horizon agent benchmarks show consistent gains under strict context limits, demonstrating that optical encoding increases effective memory capacity while preserving faithful evidence recovery.