paper-with-me

홈 › Papers

Koala: An Index for Quantifying Overlaps with Pre-training Corpora

2023-03-26 · Thuy-Trang Vu, Xuanli He, Gholamreza Haffari, Ehsan Shareghi

In very recent years more attention has been placed on probing the role of pre-training data in Large Language Models (LLMs) downstream behaviour. Despite the importance, there is no public tool that supports such analysis of pre-training corpora at large scale. To help research in this space, we launch Koala, a searchable index over large pre-training corpora using compressed suffix arrays with highly efficient compression rate and search support. In its first release we index the public proportion of OPT 175B pre-training data. Koala provides a framework to do forensic analysis on the current and future benchmarks as well as to assess the degree of memorization in the output from the LLMs. Koala is available for public use at https://koala-index.erc.monash.edu/.

📄 PDF Abstract BibTeX arXiv:2303.14770

Code (0)

등록된 구현이 없습니다.

Tasks

Memorization

Methods 이 논문이 사용한 방법론

OPT OPT is a suite of decoder-only pre-trained transformers ranging from 125M to 175B parameters. The model uses an AdamW optimizer and weight decay of 0.1. It follows a linear…

Similar Papers 제목 키워드 기반

KOALA: Enhancing Speculative Decoding for LLM via Multi-Layer Draft Heads with Adversarial Learning

2024-08-15 · Kaiqi Zhang, Jing Zhao, Rui Chen

Large Language Models (LLMs) exhibit high inference latency due to their autoregressive decoding nature. While the draft head in speculative decoding mitigates this issue, its full potential remains unexplored. In this p…

Koala-36M: A Large-scale Video Dataset Improving Consistency between Fine-grained Conditions and Video Content

2024-10-10 · CVPR 2025 1 · Qiuheng Wang, Yukai Shi, Jiarong Ou, Rui Chen 외

As visual generation technologies continue to advance, the scale of video datasets has expanded rapidly, and the quality of these datasets is critical to the performance of video generation models. We argue that temporal…

Video AlignmentVideo Generation

KoALA: KL-L0 Adversarial Detector via Label Agreement

2025-10-14 · Siqi Li, Yasser Shoukry arxiv

Deep neural networks are highly susceptible to adversarial attacks, which pose significant risks to security- and safety-critical applications. We present KoALA (KL-L0 Adversarial detection via Label Agreement), a novel,…

Adversarial Attack

KOALA++: Efficient Kalman-Based Optimization of Neural Networks with Gradient-Covariance Products

2025-06-04 · Zixuan Xia, Aram Davtyan, Paolo Favaro

We propose KOALA++, a scalable Kalman-based optimization algorithm that explicitly models structured gradient uncertainty in neural network training. Unlike second-order methods, which rely on expensive second order grad…

image-classificationImage ClassificationLanguage ModelingLanguage Modelling+1

KoALa-Bench: Evaluating Large Audio Language Models on Korean Speech Understanding and Faithfulness

2026-03-30 · Jinyoung Kim, Hyeongsoo Lim, Eunseo Seo, Minho Jang 외 arxiv

Recent advances in large audio language models (LALMs) have enabled multilingual speech understanding. However, benchmarks for evaluating LALMs remain scarce for non-English languages, with Korean being one such underexp…

Instruction FollowingSpeech RecognitionQuestion Answering