We present a baseline for the CLPsych 2025 A.1 task: classifying self-states in mental health data taken from Reddit. We use few-shot learning with a 4-bit quantized Gemma 2 9B model and a data preprocessing step which first identifies relevant sentences indicating self-state evidence, and then performs a binary classification to determine whether the sentence is evidence of an adaptive or maladaptive self-state. This system outperforms our other method which relies on an LLM to highlight spans of variable length independently. We attribute the performance of our model to the benefits of this sentence chunking step for two reasons: partitioning posts into sentences 1) broadly matches the granularity at which self-states were human-annotated and 2) simplifies the task for our language model to a binary classification problem. Our system places third out of fourteen systems submitted for Task A.1, achieving a test-time recall of 0.579.

Operational Technology Cybersecurity (OTCS) continues to be a dominant challenge for critical infrastructure such as railways. As these systems become increasingly vulnerable to malicious attacks due to digitalization, effective documentation and compliance processes are essential to protect these safety-critical systems. This paper proposes a novel system that leverages Large Language Models (LLMs) and multi-stage retrieval to enhance the compliance verification process against standards like IEC 62443 and the rail-specific IEC 63452. We first evaluate a Baseline Compliance Architecture (BCA) for answering OTCS compliance queries, then develop an extended approach called Parallel Compliance Architecture (PCA) that incorporates additional context from regulatory standards. Through empirical evaluation comparing OpenAI-gpt-4o and Claude-3.5-haiku models in these architectures, we demonstrate that the PCA significantly improves both correctness and reasoning quality in compliance verification. Our research establishes metrics for response correctness, logical reasoning, and hallucination detection, highlighting the strengths and limitations of using LLMs for compliance verification in railway cybersecurity. The results suggest that retrieval-augmented approaches can significantly improve the efficiency and accuracy of compliance assessments, particularly valuable in an industry facing a shortage of cybersecurity expertise.

As Large Language Models (LLMs) advance, their potential for widespread societal impact grows simultaneously. Hence, rigorous LLM evaluations are both a technical necessity and social imperative. While numerous evaluation benchmarks have been developed, there remains a critical gap in meta-evaluation: effectively assessing benchmarks' quality. We propose MEQA, a framework for the meta-evaluation of question and answer (QA) benchmarks, to provide standardized assessments, quantifiable scores, and enable meaningful intra-benchmark comparisons. We demonstrate this approach on cybersecurity benchmarks, using human and LLM evaluators, highlighting the benchmarks' strengths and weaknesses. We motivate our choice of test domain by AI models' dual nature as powerful defensive tools and security threats.

Conducting data analysis typically involves authoring code to transform, visualize, analyze, and interpret data. Large language models (LLMs) are now capable of generating such code for simple, routine analyses. LLMs promise to democratize data science by enabling those with limited programming expertise to conduct data analyses, including in scientific research, business, and policymaking. However, analysts in many real-world settings must often exercise fine-grained control over specific analysis steps, verify intermediate results explicitly, and iteratively refine their analytical approaches. Such tasks present barriers to building robust and reproducible analyses using LLMs alone or even in conjunction with existing authoring tools (e.g., computational notebooks). This paper introduces Flowco, a new mixed-initiative system to address these challenges. Flowco leverages a visual dataflow programming model and integrates LLMs into every phase of the authoring process. A user study suggests that Flowco supports analysts, particularly those with less programming experience, in quickly authoring, debugging, and refining data analyses.

In recent years, the fashion industry has increasingly adopted AI technologies to enhance customer experience, driven by the proliferation of e-commerce platforms and virtual applications. Among the various tasks, virtual try-on and multimodal fashion image editing -- which utilizes diverse input modalities such as text, garment sketches, and body poses -- have become a key area of research. Diffusion models have emerged as a leading approach for such generative tasks, offering superior image quality and diversity. However, most existing virtual try-on methods rely on having a specific garment input, which is often impractical in real-world scenarios where users may only provide textual specifications. To address this limitation, in this work we introduce Fashion Retrieval-Augmented Generation (Fashion-RAG), a novel method that enables the customization of fashion items based on user preferences provided in textual form. Our approach retrieves multiple garments that match the input specifications and generates a personalized image by incorporating attributes from the retrieved items. To achieve this, we employ textual inversion techniques, where retrieved garment images are projected into the textual embedding space of the Stable Diffusion text encoder, allowing seamless integration of retrieved elements into the generative process. Experimental results on the Dress Code dataset demonstrate that Fashion-RAG outperforms existing methods both qualitatively and quantitatively, effectively capturing fine-grained visual details from retrieved garments. To the best of our knowledge, this is the first work to introduce a retrieval-augmented generation approach specifically tailored for multimodal fashion image editing.

--Consumers often heavily rely on online product reviews, analyzing both quantitative ratings and textual descriptions to assess product quality. However, existing research hasn't adequately addressed how to systematically encourage the creation of comprehensive reviews that capture both customers sentiment and detailed product feature analysis. This paper presents CPR, a novel methodology that leverages the power of Large Language Models (LLMs) and T opic Modeling to guide users in crafting insightful and well-rounded reviews. Our approach employs a three-stage process: first, we present users with product-specific terms for rating; second, we generate targeted phrase suggestions based on these ratings; and third, we integrate user-written text through topic modeling, ensuring all key aspects are addressed. We evaluate CPR using text-to-text LLMs, comparing its performance against real-world customer reviews from Walmart. Our results demonstrate that CPR effectively identifies relevant product terms, even for new products lacking prior reviews, and provides sentiment-aligned phrase suggestions, saving users time and enhancing reviews quality. Quantitative analysis reveals a 12.3% improvement in BLEU score over baseline methods, further supported by manual evaluation of generated phrases. We conclude by discussing potential extensions and future research directions. I NTRODUCTION Product reviews play a crucial role for retailers, as they help build trust among potential customers by providing social proof. They influence purchase decisions [7], [9], [19], [25] by offering information on the quality and suitability of the product. Reviews also provide valuable feedback for retailers, allows them to improve their products and enhance customer satisfaction. Furthermore, product reviews contribute to product search optimization efforts [8], giving retailers a competitive advantage and fostering customer engagement and loyalty. Product review phrase suggestion is a sub-task of text-to-text generation in natural language processing (NLP). Online shopping is increasingly popular. However, customers often lack the motivation to write constructive reviews.

This work defends the 'Whole Hog Thesis': sophisticated Large Language Models (LLMs) like ChatGPT are full-blown linguistic and cognitive agents, possessing understanding, beliefs, desires, knowledge, and intentions. We argue against prevailing methodologies in AI philosophy, rejecting starting points based on low-level computational details ('Just an X' fallacy) or pre-existing theories of mind. Instead, we advocate starting with simple, high-level observations of LLM behavior (e.g., answering questions, making suggestions) -- defending this data against charges of metaphor, loose talk, or pretense. From these observations, we employ 'Holistic Network Assumptions' -- plausible connections between mental capacities (e.g., answering implies knowledge, knowledge implies belief, action implies intention) -- to argue for the full suite of cognitive states. We systematically rebut objections based on LLM failures (hallucinations, planning/reasoning errors), arguing these don't preclude agency, often mirroring human fallibility. We address numerous 'Games of Lacks', arguing that LLMs do not lack purported necessary conditions for cognition (e.g., semantic grounding, embodiment, justification, intrinsic intentionality) or that these conditions are not truly necessary, often relying on anti-discriminatory arguments comparing LLMs to diverse human capacities. Our approach is evidential, not functionalist, and deliberately excludes consciousness. We conclude by speculating on the possibility of LLMs possessing 'alien' contents beyond human conceptual schemes.

To reduce the time and computational costs of inference of large language models, there has been interest in parameter-efficient low-rank early-exit casting of transformer hidden-representations to final-representations. Such low-rank short-cutting has been shown to outperform identity shortcuts at early model stages while offering parameter-efficiency in shortcut jumps. However, current low-rank methods maintain a separate early-exit shortcut jump to final-representations for each transformer intermediate block-level during inference. In this work, we propose selection of a single One-Jump-Fits-All (OJFA) low-rank shortcut that offers over a 30x reduction in shortcut parameter costs during inference. We show that despite this extreme reduction, our OJFA choice largely matches the performance of maintaining multiple shortcut jumps during inference and offers stable precision from all transformer block-levels for GPT2-XL, Phi3-Mini and Llama2-7B transformer models.

Efficiently handling long contexts in transformer-based language models with low perplexity is an active area of research. Numerous recent approaches like Linformer, Longformer, Performer, and Structured state space models (SSMs)., have not fully resolved this problem. All these models strive to reduce the quadratic time complexity of the attention mechanism while minimizing the loss in quality due to the effective compression of the long context. Inspired by the cache and virtual memory principle in computers, where in case of a cache miss, not only the needed data is retrieved from the memory, but the adjacent data is also obtained, we apply this concept to handling long contexts by dividing it into small segments. In our design, we retrieve the nearby segments in an uncompressed form when high segment-level attention occurs at the compressed level. Our en-hancements for handling long context include aggregating four attention mechanisms consisting of short sliding window attention, long compressed segmented attention, dynamically retrieving top k high attention uncompressed segments, and overlapping segments in long segment attention to avoid segment fragmentation. These enhancements result in an architecture that outperforms ex-isting SOTA architectures with an average perplexity improvement of 8.5% over similar model sizes.

Reinforcement Learning from Human Feedback (RLHF) is central in aligning large language models (LLMs) with human values and expectations. However, the process remains susceptible to governance challenges, including evaluator bias, inconsistency, and the unreliability of feedback. This study examines how the cognitive capacity of evaluators, specifically their level of rationality, affects the stability of reinforcement signals. A controlled experiment comparing high-rationality and low-rationality participants reveals that evaluators with higher rationality scores produce significantly more consistent and expert-aligned feedback. In contrast, lower-rationality participants demonstrate considerable variability in their reinforcement decisions ($p < 0.01$). To address these challenges and improve RLHF governance, we recommend implementing evaluator pre-screening, systematic auditing of feedback consistency, and reliability-weighted reinforcement aggregation. These measures enhance the fairness, transparency, and robustness of AI alignment pipelines.