Large Language Models (LLMs) are leading a new technological revolution as one of the most promising research streams toward artificial general intelligence. The scaling of these models, accomplished by increasing the number of parameters and the magnitude of the training datasets, has been linked to various so-called emergent abilities that were previously unobserved. These emergent abilities, ranging from advanced reasoning and in-context learning to coding and problem-solving, have sparked an intense scientific debate: Are they truly emergent, or do they simply depend on external factors, such as training dynamics, the type of problems, or the chosen metric? What underlying mechanism causes them? Despite their transformative potential, emergent abilities remain poorly understood, leading to misconceptions about their definition, nature, predictability, and implications. In this work, we shed light on emergent abilities by conducting a comprehensive review of the phenomenon, addressing both its scientific underpinnings and real-world consequences. We first critically analyze existing definitions, exposing inconsistencies in conceptualizing emergent abilities. We then explore the conditions under which these abilities appear, evaluating the role of scaling laws, task complexity, pre-training loss, quantization, and prompting strategies. Our review extends beyond traditional LLMs and includes Large Reasoning Models (LRMs), which leverage reinforcement learning and inference-time search to amplify reasoning and self-reflection. However, emergence is not inherently positive. As AI systems gain autonomous reasoning capabilities, they also develop harmful behaviors, including deception, manipulation, and reward hacking. We highlight growing concerns about safety and governance, emphasizing the need for better evaluation frameworks and regulatory oversight.
With the advancement of conversational large language models (LLMs), several LLM-based Conversational Shopping Agents (CSA) have been developed to help customers smooth their online shopping. The primary objective in building an engaging and trustworthy CSA is to ensure the agent's responses about product factoids are accurate and factually grounded. However, two challenges remain. First, LLMs produce hallucinated or unsupported claims. Such inaccuracies risk spreading misinformation and diminishing customer trust. Second, without providing knowledge source attribution in CSA response, customers struggle to verify LLM-generated information. To address both challenges, we present an easily productionized solution that enables a ''citation experience'' to our customers. We build auto-evaluation metrics to holistically evaluate LLM's grounding and attribution capabilities, suggesting that citation generation paradigm substantially improves grounding performance by 13.83%. To deploy this capability at scale, we introduce Multi-UX-Inference system, which appends source citations to LLM outputs while preserving existing user experience features and supporting scalable inference. Large-scale online A/B tests show that grounded CSA responses improves customer engagement by 3% - 10%, depending on UX variations.
As large language models (LLMs) gain widespread attention in both academia and industry, it becomes increasingly critical and challenging to effectively evaluate their capabilities. Existing evaluation methods can be broadly categorized into two types: manual evaluation and automatic evaluation. Manual evaluation, while comprehensive, is often costly and resource-intensive. Conversely, automatic evaluation offers greater scalability but is constrained by the limitations of its evaluation criteria (dominated by reference-based answers). To address these challenges, NTCIR-18 introduced the AEOLLM (Automatic Evaluation of LLMs) task, aiming to encourage reference-free evaluation methods that can overcome the limitations of existing approaches. In this paper, to enhance the evaluation performance of the AEOLLM task, we propose three key methods to improve the reference-free evaluation: 1) Multi-model Collaboration: Leveraging multiple LLMs to approximate human ratings across various subtasks; 2) Prompt Auto-optimization: Utilizing LLMs to iteratively refine the initial task prompts based on evaluation feedback from training samples; and 3) In-context Learning (ICL) Optimization: Based on the multi-task evaluation feedback, we train a specialized in-context example retrieval model, combined with a semantic relevance retrieval model, to jointly identify the most effective in-context learning examples. Experiments conducted on the final dataset demonstrate that our approach achieves superior performance on the AEOLLM task.
Analysis of parliamentary speeches and political-party manifestos has become an integral area of computational study of political texts. While speeches have been overwhelmingly analysed using unsupervised methods, a large corpus of manifestos with by-statement political-stance labels has been created by the participants of the MARPOR project. It has been recently shown that these labels can be predicted by a neural model; however, the current approach relies on provided statement boundaries, limiting out-of-domain applicability. In this work, we propose and test a range of unified split-and-label frameworks -- based on linear-chain CRFs, fine-tuned text-to-text models, and the combination of in-context learning with constrained decoding -- that can be used to jointly segment and classify statements from raw textual data. We show that our approaches achieve competitive accuracy when applied to raw text of political manifestos, and then demonstrate the research potential of our method by applying it to the records of the UK House of Commons and tracing the political trajectories of four major parties in the last three decades.
Video-based dialogue systems, such as education assistants, have compelling application value, thereby garnering growing interest. However, the current video-based dialogue systems are limited by their reliance on a single dialogue type, which hinders their versatility in practical applications across a range of scenarios, including question-answering, emotional dialog, etc. In this paper, we identify this challenge as how to generate video-driven multilingual mixed-type dialogues. To mitigate this challenge, we propose a novel task and create a human-to-human video-driven multilingual mixed-type dialogue corpus, termed KwaiChat, containing a total of 93,209 videos and 246,080 dialogues, across 4 dialogue types, 30 domains, 4 languages, and 13 topics. Additionally, we establish baseline models on KwaiChat. An extensive analysis of 7 distinct LLMs on KwaiChat reveals that GPT-4o achieves the best performance but still cannot perform well in this situation even with the help of in-context learning and fine-tuning, which indicates that the task is not trivial and needs further research.
Conversational Recommender Systems (CRSs) have emerged as a transformative paradigm for offering personalized recommendations through natural language dialogue. However, they face challenges with knowledge sparsity, as users often provide brief, incomplete preference statements. While recent methods have integrated external knowledge sources to mitigate this, they still struggle with semantic understanding and complex preference reasoning. Recent Large Language Models (LLMs) demonstrate promising capabilities in natural language understanding and reasoning, showing significant potential for CRSs. Nevertheless, due to the lack of domain knowledge, existing LLM-based CRSs either produce hallucinated recommendations or demand expensive domain-specific training, which largely limits their applicability. In this work, we present G-CRS (Graph Retrieval-Augmented Large Language Model for Conversational Recommender Systems), a novel training-free framework that combines graph retrieval-augmented generation and in-context learning to enhance LLMs' recommendation capabilities. Specifically, G-CRS employs a two-stage retrieve-and-recommend architecture, where a GNN-based graph reasoner first identifies candidate items, followed by Personalized PageRank exploration to jointly discover potential items and similar user interactions. These retrieved contexts are then transformed into structured prompts for LLM reasoning, enabling contextually grounded recommendations without task-specific training. Extensive experiments on two public datasets show that G-CRS achieves superior recommendation performance compared to existing methods without requiring task-specific training.
Finetuning provides a scalable and cost-effective means of customizing language models for specific tasks or response styles, with greater reliability than prompting or in-context learning. In contrast, the conventional wisdom is that injecting knowledge via finetuning results in brittle performance and poor generalization. We argue that the dichotomy of "task customization" (e.g., instruction tuning) and "knowledge injection" (e.g., teaching new facts) is a distinction without a difference. We instead identify concrete factors that explain the heterogeneous effectiveness observed with finetuning. To this end, we conduct a large-scale experimental study of finetuning the frontier Gemini v1.5 model family on a spectrum of datasets that are artificially engineered to interpolate between the strengths and failure modes of finetuning. Our findings indicate that question-answer training data formats provide much stronger knowledge generalization than document/article-style training data, numerical information can be harder for finetuning to retain than categorical information, and models struggle to apply finetuned knowledge during multi-step reasoning even when trained on similar examples -- all factors that render "knowledge injection" to be especially difficult, even after controlling for considerations like data augmentation and information volume. On the other hand, our findings also indicate that it is not fundamentally more difficult to finetune information about a real-world event than information about what a model's writing style should be.
Recently, large language models (LLMs) with hundreds of billions of parameters have demonstrated the emergent ability, surpassing traditional methods in various domains even without fine-tuning over domain-specific data. However, when it comes to financial sentiment analysis (FSA)$\unicode{x2013}$a fundamental task in financial AI$\unicode{x2013}$these models often encounter various challenges, such as complex financial terminology, subjective human emotions, and ambiguous inclination expressions. In this paper, we aim to answer the fundamental question: whether LLMs are good in-context learners for FSA? Unveiling this question can yield informative insights on whether LLMs can learn to address the challenges by generalizing in-context demonstrations of financial document-sentiment pairs to the sentiment analysis of new documents, given that finetuning these models on finance-specific data is difficult, if not impossible at all. To the best of our knowledge, this is the first paper exploring in-context learning for FSA that covers most modern LLMs (recently released DeepSeek V3 included) and multiple in-context sample selection methods. Comprehensive experiments validate the in-context learning capability of LLMs for FSA.
As large language models (LLMs) gradually become integral tools for problem solving in daily life worldwide, understanding linguistic inequality is becoming increasingly important. Existing research has primarily focused on static analyses that assess the disparities in the existing knowledge and capabilities of LLMs across languages. However, LLMs are continuously evolving, acquiring new knowledge to generate up-to-date, domain-specific responses. Investigating linguistic inequalities within this dynamic process is, therefore, also essential. In this paper, we explore inequalities in new knowledge learning by LLMs across different languages and four key dimensions: effectiveness, transferability, prioritization, and robustness. Through extensive experiments under two settings (in-context learning and fine-tuning) using both proprietary and open-source models, we demonstrate that low-resource languages consistently face disadvantages across all four dimensions. By shedding light on these disparities, we aim to raise awareness of linguistic inequalities in LLMs' new knowledge learning, fostering the development of more inclusive and equitable future LLMs.
Understanding the semantics of columns in relational tables is an important pre-processing step for indexing data lakes in order to provide rich data search. An approach to establishing such understanding is column type annotation (CTA) where the goal is to annotate table columns with terms from a given vocabulary. This paper experimentally compares different knowledge generation and self-refinement strategies for LLM-based column type annotation. The strategies include using LLMs to generate term definitions, error-based refinement of term definitions, self-correction, and fine-tuning using examples and term definitions. We evaluate these strategies along two dimensions: effectiveness measured as F1 performance and efficiency measured in terms of token usage and cost. Our experiments show that the best performing strategy depends on the model/dataset combination. We find that using training data to generate label definitions outperforms using the same data as demonstrations for in-context learning for two out of three datasets using OpenAI models. The experiments further show that using the LLMs to refine label definitions brings an average increase of 3.9% F1 in 10 out of 12 setups compared to the performance of the non-refined definitions. Combining fine-tuned models with self-refined term definitions results in the overall highest performance, outperforming zero-shot prompting fine-tuned models by at least 3% in F1 score. The costs analysis shows that while reaching similar F1 score, self-refinement via prompting is more cost efficient for use cases requiring smaller amounts of tables to be annotated while fine-tuning is more efficient for large amounts of tables.
Existing domain-specific Large Language Models (LLMs) are typically developed by fine-tuning general-purposed LLMs with large-scale domain-specific corpora. However, training on large-scale corpora often fails to effectively organize domain knowledge of LLMs, leading to fragmented understanding. Inspired by how humans connect concepts and organize knowledge through mind maps, we aim to emulate this approach by using ontology with hierarchical conceptual knowledge to reorganize LLM's domain knowledge. From this perspective, we propose an ontology-driven self-training framework called OntoTune, which aims to align LLMs with ontology through in-context learning, enabling the generation of responses guided by the ontology. We leverage in-context learning to identify whether the LLM has acquired the specific concept's ontology knowledge, and select the entries not yet mastered by LLM as the training set to further align the LLM with ontology. Compared to existing domain LLMs based on newly collected large-scale domain-specific corpora, our OntoTune, which relies on the existing, long-term developed ontology and LLM itself, significantly reduces data maintenance costs and offers improved generalization ability. We conduct our study in the medical domain to evaluate the effectiveness of OntoTune, utilizing a standardized medical ontology, SNOMED CT as our ontology source. Experimental results demonstrate that OntoTune achieves state-of-the-art performance in both in-ontology task hypernym discovery and out-of-ontology task medical domain QA. Moreover, compared to the latest direct ontology injection method TaxoLLaMA, our OntoTune better preserves original knowledge of LLM. The code and data are available at https://github.com/zjukg/OntoTune.
The in-context learning capabilities of large language models (LLMs) show great potential in mental health support. However, the lack of counseling datasets, particularly in Chinese corpora, restricts their application in this field. To address this, we constructed Psy-Insight, the first mental health-oriented explainable multi-task bilingual dataset. We collected face-to-face multi-turn counseling dialogues, which are annotated with multi-task labels and conversation process explanations. Our annotations include psychotherapy, emotion, strategy, and topic labels, as well as turn-level reasoning and session-level guidance. Psy-Insight is not only suitable for tasks such as label recognition but also meets the need for training LLMs to act as empathetic counselors through logical reasoning. Experiments show that training LLMs on Psy-Insight enables the models to not only mimic the conversation style but also understand the underlying strategies and reasoning of counseling.
Combinatorial optimization problems often rely on heuristic algorithms to generate efficient solutions. However, the manual design of heuristics is resource-intensive and constrained by the designer's expertise. Recent advances in artificial intelligence, particularly large language models (LLMs), have demonstrated the potential to automate heuristic generation through evolutionary frameworks. Recent works focus only on well-known combinatorial optimization problems like the traveling salesman problem and online bin packing problem when designing constructive heuristics. This study investigates whether LLMs can effectively generate heuristics for niche, not yet broadly researched optimization problems, using the unit-load pre-marshalling problem as an example case. We propose the Contextual Evolution of Heuristics (CEoH) framework, an extension of the Evolution of Heuristics (EoH) framework, which incorporates problem-specific descriptions to enhance in-context learning during heuristic generation. Through computational experiments, we evaluate CEoH and EoH and compare the results. Results indicate that CEoH enables smaller LLMs to generate high-quality heuristics more consistently and even outperform larger models. Larger models demonstrate robust performance with or without contextualized prompts. The generated heuristics exhibit scalability to diverse instance configurations.
Multiple previous studies have reported suboptimal performance of LLMs in biomedical text mining. By analyzing failure patterns in these evaluations, we identified three primary challenges for LLMs in biomedical corpora: (1) LLMs fail to learn implicit dataset-specific nuances from supervised data, (2) The common formatting requirements of discriminative tasks limit the reasoning capabilities of LLMs particularly for LLMs that lack test-time compute, and (3) LLMs struggle to adhere to annotation guidelines and match exact schemas, which hinders their ability to understand detailed annotation requirements which is essential in biomedical annotation workflow. We experimented with prompt engineering techniques targeted to the above issues, and developed a pipeline that dynamically extracts instructions from annotation guidelines. Our results show that frontier LLMs can approach or surpass the performance of SOTA BERT-based models with minimal reliance on manually annotated data and without fine-tuning. Furthermore, we performed model distillation on a closed-source LLM, demonstrating that a BERT model trained exclusively on synthetic data annotated by LLMs can also achieve a practical performance. Based on these findings, we explored the feasibility of partially replacing manual annotation with LLMs in production scenarios for biomedical text mining.
This paper presents a compact model that achieves strong sentiment analysis capabilities through targeted distillation from advanced large language models (LLMs). Our methodology decouples the distillation target into two key components: sentiment-related knowledge and task alignment. To transfer these components, we propose a two-stage distillation framework. The first stage, knowledge-driven distillation (\textsc{KnowDist}), transfers sentiment-related knowledge to enhance fundamental sentiment analysis capabilities. The second stage, in-context learning distillation (\textsc{ICLDist}), transfers task-specific prompt-following abilities to optimize task alignment. For evaluation, we introduce \textsc{SentiBench}, a comprehensive sentiment analysis benchmark comprising 3 task categories across 12 datasets. Experiments on this benchmark demonstrate that our model effectively balances model size and performance, showing strong competitiveness compared to existing small-scale LLMs.
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.
The high cost of obtaining high-quality annotated data for in-context learning (ICL) has motivated the development of methods that use self-generated annotations in place of ground-truth labels. While these approaches have shown promising results in few-shot settings, they generally do not scale to many-shot scenarios. In this work, we study ICL with self-generated examples using a framework analogous to traditional semi-supervised learning, consisting of annotation generation, demonstration selection, and in-context inference. Within this framework, we propose a simple baseline that outperforms ground-truth ICL in zero-shot, few-shot, and many-shot settings. Notably, we observe a scaling law with this baseline, where optimal performance is achieved with more than 1,000 demonstrations. To fully exploit the many-shot capabilities of semi-supervised ICL, we introduce IterPSD, an iterative annotation approach that integrates iterative refinement and curriculum pseudo-labeling techniques from semi-supervised learning, yielding up to 6.8% additional gains on classification tasks.
In-context learning (ICL), a type of universal model, demonstrates exceptional generalization across a wide range of tasks without retraining by leveraging task-specific guidance from context, making it particularly effective for the complex demands of neuroimaging. However, existing ICL models, which take 2D images as input, struggle to fully leverage the 3D anatomical structures in neuroimages, leading to a lack of global awareness and suboptimal performance. In this regard, we introduce Neuroverse3D, an ICL model capable of performing multiple neuroimaging tasks (e.g., segmentation, denoising, inpainting) in 3D. Neuroverse3D overcomes the large memory consumption due to 3D inputs through adaptive parallel-sequential context processing and a U-shape fusion strategy, allowing it to handle an unlimited number of context images. Additionally, we propose an optimized loss to balance multi-task training and enhance the focus on anatomical structures. Our study incorporates 43,674 3D scans from 19 neuroimaging datasets and evaluates Neuroverse3D on 14 diverse tasks using held-out test sets. The results demonstrate that Neuroverse3D significantly outperforms existing ICL models and closely matches the performance of task-specific models. The code and model weights are publicly released at: https://github.com/jiesihu/Neu3D.
Evolutionary Algorithms (EAs) employ random or simplistic selection methods, limiting their exploration of solution spaces and convergence to optimal solutions. The randomness in performing crossover or mutations may limit the model's ability to evolve efficiently. This paper introduces Preference-Aligned Individual Reciprocity (PAIR), a novel selection approach leveraging Large Language Models to emulate human-like mate selection, thereby introducing intelligence to the pairing process in EAs. PAIR prompts an LLM to evaluate individuals within a population based on genetic diversity, fitness level, and crossover compatibility, guiding more informed pairing decisions. We evaluated PAIR against a baseline method called LLM-driven EA (LMEA), published recently. Results indicate that PAIR significantly outperforms LMEA across various TSP instances, achieving lower optimality gaps and improved convergence. This performance is especially noticeable when combined with the flash thinking model, demonstrating increased population diversity to escape local optima. In general, PAIR provides a new strategy in the area of in-context learning for LLM-driven selection in EAs via sophisticated preference modelling, paving the way for improved solutions and further studies into LLM-guided optimization.
The in-context learning capabilities of modern language models have motivated a deeper mathematical understanding of sequence models. A line of recent work has shown that linear attention models can emulate projected gradient descent iterations to implicitly learn the task vector from the data provided in the context window. In this work, we consider a novel setting where the global task distribution can be partitioned into a union of conditional task distributions. We then examine the use of task-specific prompts and prediction heads for learning the prior information associated with the conditional task distribution using a one-layer attention model. Our results on loss landscape show that task-specific prompts facilitate a covariance-mean decoupling where prompt-tuning explains the conditional mean of the distribution whereas the variance is learned/explained through in-context learning. Incorporating task-specific head further aids this process by entirely decoupling estimation of mean and variance components. This covariance-mean perspective similarly explains how jointly training prompt and attention weights can provably help over fine-tuning after pretraining.