Search Results for author:
Found 48 papers, 37 papers with code
Precise Localization of Memories: A Fine-grained Neuron-level Knowledge Editing Technique for LLMs
Knowledge editing aims to update outdated information in Large Language Models (LLMs).
Sparse Auto-Encoder Interprets Linguistic Features in Large Language Models
Large language models (LLMs) excel in tasks that require complex linguistic abilities, such as reference disambiguation and metaphor recognition/generation.
Agentic Reward Modeling: Integrating Human Preferences with Verifiable Correctness Signals for Reliable Reward Systems
In this paper, we propose agentic reward modeling, a reward system that combines reward models with verifiable correctness signals from different aspects to provide reliable rewards.
LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitasks
This paper introduces LongBench v2, a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks.
Constraint Back-translation Improves Complex Instruction Following of Large Language Models
Following the conventional instruction-tuning practice, previous works conduct post-training on complex instruction-response pairs generated by feeding complex instructions to advanced LLMs.
Configurable Foundation Models: Building LLMs from a Modular Perspective
We first formalize modules into emergent bricks - functional neuron partitions that emerge during the pre-training phase, and customized bricks - bricks constructed via additional post-training to improve the capabilities and knowledge of LLMs.
OpenEP: Open-Ended Future Event Prediction
For question construction, we pose questions from seven perspectives, including location, time, event development, event outcome, event impact, event response, and other, to facilitate an in-depth analysis and understanding of the comprehensive evolution of events.
MAVEN-Fact: A Large-scale Event Factuality Detection Dataset
Thanks to the comprehensive annotations of event arguments and relations in MAVEN, MAVEN-Fact also supports some further analyses and we find that adopting event arguments and relations helps in event factuality detection for fine-tuned models but does not benefit LLMs.
Finding Safety Neurons in Large Language Models
In this paper, we explore the inner mechanisms of safety alignment from the perspective of mechanistic interpretability, focusing on identifying and analyzing safety neurons within LLMs that are responsible for safety behaviors.
R-Eval: A Unified Toolkit for Evaluating Domain Knowledge of Retrieval Augmented Large Language Models
In this paper, we address the challenges of evaluating RALLMs by introducing the R-Eval toolkit, a Python toolkit designed to streamline the evaluation of different RAG workflows in conjunction with LLMs.
Event GDR: Event-Centric Generative Document Retrieval
Inspired by this, we propose Event GDR, an event-centric generative document retrieval model, integrating event knowledge into this task.
TacoERE: Cluster-aware Compression for Event Relation Extraction
Existing work mainly focuses on directly modeling the entire document, which cannot effectively handle long-range dependencies and information redundancy.
ADELIE: Aligning Large Language Models on Information Extraction
Large language models (LLMs) usually fall short on information extraction (IE) tasks and struggle to follow the complex instructions of IE tasks.
Event-level Knowledge Editing
However, natural knowledge updates in the real world come from the occurrences of new events rather than direct changes in factual triplets.
When does In-context Learning Fall Short and Why? A Study on Specification-Heavy Tasks
In this paper, we find that ICL falls short of handling specification-heavy tasks, which are tasks with complicated and extensive task specifications, requiring several hours for ordinary humans to master, such as traditional information extraction tasks.
MAVEN-Arg: Completing the Puzzle of All-in-One Event Understanding Dataset with Event Argument Annotation
Understanding events in texts is a core objective of natural language understanding, which requires detecting event occurrences, extracting event arguments, and analyzing inter-event relationships.
Finding and Editing Multi-Modal Neurons in Pre-Trained Transformers
Understanding the internal mechanisms by which multi-modal large language models (LLMs) interpret different modalities and integrate cross-modal representations is becoming increasingly critical for continuous improvements in both academia and industry.
Mastering the Task of Open Information Extraction with Large Language Models and Consistent Reasoning Environment
Open Information Extraction (OIE) aims to extract objective structured knowledge from natural texts, which has attracted growing attention to build dedicated models with human experience.
OmniEvent: A Comprehensive, Fair, and Easy-to-Use Toolkit for Event Understanding
Event understanding aims at understanding the content and relationship of events within texts, which covers multiple complicated information extraction tasks: event detection, event argument extraction, and event relation extraction.
KoLA: Carefully Benchmarking World Knowledge of Large Language Models
The unprecedented performance of large language models (LLMs) necessitates improvements in evaluations.
The Devil is in the Details: On the Pitfalls of Event Extraction Evaluation
In this paper, we check the reliability of EE evaluations and identify three major pitfalls: (1) The data preprocessing discrepancy makes the evaluation results on the same dataset not directly comparable, but the data preprocessing details are not widely noted and specified in papers.
Emergent Modularity in Pre-trained Transformers
In analogy to human brains, we consider two main characteristics of modularity: (1) functional specialization of neurons: we evaluate whether each neuron is mainly specialized in a certain function, and find that the answer is yes.
Preserving Knowledge Invariance: Rethinking Robustness Evaluation of Open Information Extraction
In this paper, we present the first benchmark that simulates the evaluation of open information extraction models in the real world, where the syntactic and expressive distributions under the same knowledge meaning may drift variously.
ChatLog: Carefully Evaluating the Evolution of ChatGPT Across Time
We find some stable features that stay unchanged and apply them on the detection of ChatGPT-generated texts to improve the robustness of cross-version detection.
GOAL: A Challenging Knowledge-grounded Video Captioning Benchmark for Real-time Soccer Commentary Generation
Despite the recent emergence of video captioning models, how to generate vivid, fine-grained video descriptions based on the background knowledge (i. e., long and informative commentary about the domain-specific scenes with appropriate reasoning) is still far from being solved, which however has great applications such as automatic sports narrative.
Language-Specific Representation of Emotion-Concept Knowledge Causally Supports Emotion Inference
Humans no doubt use language to communicate about their emotional experiences, but does language in turn help humans understand emotions, or is language just a vehicle of communication?
READIN: A Chinese Multi-Task Benchmark with Realistic and Diverse Input Noises
In order to fill this important gap, we construct READIN: a Chinese multi-task benchmark with REalistic And Diverse Input Noises.
MAVEN-ERE: A Unified Large-scale Dataset for Event Coreference, Temporal, Causal, and Subevent Relation Extraction
It contains 103, 193 event coreference chains, 1, 216, 217 temporal relations, 57, 992 causal relations, and 15, 841 subevent relations, which is larger than existing datasets of all the ERE tasks by at least an order of magnitude.
Finding Skill Neurons in Pre-trained Transformer-based Language Models
Furthermore, we demonstrate the skill neurons are most likely generated in pre-training rather than fine-tuning by showing that the skill neurons found with prompt tuning are also crucial for other fine-tuning methods freezing neuron weights, such as the adapter-based tuning and BitFit.
COPEN: Probing Conceptual Knowledge in Pre-trained Language Models
We believe this is a critical bottleneck for realizing human-like cognition in PLMs.
A Roadmap for Big Model
With the rapid development of deep learning, training Big Models (BMs) for multiple downstream tasks becomes a popular paradigm.
LEVEN: A Large-Scale Chinese Legal Event Detection Dataset
However, existing Legal Event Detection (LED) datasets only concern incomprehensive event types and have limited annotated data, which restricts the development of LED methods and their downstream applications.
Delta Tuning: A Comprehensive Study of Parameter Efficient Methods for Pre-trained Language Models
This necessitates a new branch of research focusing on the parameter-efficient adaptation of PLMs, dubbed as delta tuning in this paper.
CUGE: A Chinese Language Understanding and Generation Evaluation Benchmark
We argue that for general-purpose language intelligence evaluation, the benchmark itself needs to be comprehensive and systematic.
Doppler velocity-based algorithm for Clustering and Velocity Estimation of moving objects
Then we estimate the velocity of the moving objects using the estimated LiDAR velocity and the Doppler velocity of moving objects obtained by clustering.
On Transferability of Prompt Tuning for Natural Language Processing
To explore whether we can improve PT via prompt transfer, we empirically investigate the transferability of soft prompts across different downstream tasks and PLMs in this work.
Exploring Universal Intrinsic Task Subspace via Prompt Tuning
In the experiments, we study diverse few-shot NLP tasks and surprisingly find that in a 250-dimensional subspace found with 100 tasks, by only tuning 250 free parameters, we can recover 97% and 83% of the full prompt tuning performance for 100 seen tasks (using different training data) and 20 unseen tasks, respectively, showing great generalization ability of the found intrinsic task subspace.
Sub-Character Tokenization for Chinese Pretrained Language Models
2) Pronunciation-based SubChar tokenizers can encode Chinese homophones into the same transliteration sequences and produce the same tokenization output, hence being robust to homophone typos.
CLEVE: Contrastive Pre-training for Event Extraction
Event extraction (EE) has considerably benefited from pre-trained language models (PLMs) by fine-tuning.
Neural Gibbs Sampling for Joint Event Argument Extraction
Existing EAE methods either extract each event argument roles independently or sequentially, which cannot adequately model the joint probability distribution among event arguments and their roles.
CPM: A Large-scale Generative Chinese Pre-trained Language Model
However, applying GPT-3 to address Chinese NLP tasks is still challenging, as the training corpus of GPT-3 is primarily English, and the parameters are not publicly available.
MAVEN: A Massive General Domain Event Detection Dataset
Most existing datasets exhibit the following issues that limit further development of ED: (1) Data scarcity.
Train No Evil: Selective Masking for Task-Guided Pre-Training
In this stage, the model is trained by masked language modeling on in-domain unsupervised data to learn domain-specific patterns and we propose a novel selective masking strategy to learn task-specific patterns.
KEPLER: A Unified Model for Knowledge Embedding and Pre-trained Language Representation
Pre-trained language representation models (PLMs) cannot well capture factual knowledge from text.
Adversarial Language Games for Advanced Natural Language Intelligence
In this work, we propose a challenging adversarial language game called Adversarial Taboo as an example, in which an attacker and a defender compete around a target word.
HMEAE: Hierarchical Modular Event Argument Extraction
Existing event extraction methods classify each argument role independently, ignoring the conceptual correlations between different argument roles.
Adversarial Training for Weakly Supervised Event Detection
Modern weakly supervised methods for event detection (ED) avoid time-consuming human annotation and achieve promising results by learning from auto-labeled data.
Adversarial Multi-lingual Neural Relation Extraction
To address these issues, we propose an adversarial multi-lingual neural relation extraction (AMNRE) model, which builds both consistent and individual representations for each sentence to consider the consistency and diversity among languages.
