English
Related papers

Related papers: Dynamically Anchored Prompting for Task-Imbalanced…

200 papers

Learning from non-stationary data streams, also called Task-Free Continual Learning (TFCL) remains challenging due to the absence of explicit task information. Although recently some methods have been proposed for TFCL, they lack…

Computer Vision and Pattern Recognition · Computer Science 2022-10-14 Fei Ye , Adrian G. Bors

In this paper, we present our finding that prepending a Task-Agnostic Prefix Prompt (TAPP) to the input improves the instruction-following ability of various Large Language Models (LLMs) during inference. TAPP is different from canonical…

Computation and Language · Computer Science 2023-12-27 Seonghyeon Ye , Hyeonbin Hwang , Sohee Yang , Hyeongu Yun , Yireun Kim , Minjoon Seo

Despite the recent progress in incremental learning, addressing catastrophic forgetting under distributional drift is still an open and important problem. Indeed, while state-of-the-art domain incremental learning (DIL) methods perform…

Computer Vision and Pattern Recognition · Computer Science 2024-11-06 Julien Nicolas , Florent Chiaroni , Imtiaz Ziko , Ola Ahmad , Christian Desrosiers , Jose Dolz

While continuous diffusion models have achieved remarkable success, discrete diffusion offers a unified framework for jointly modeling text and images. Beyond unification, discrete diffusion provides faster inference, finer control, and…

Domain incremental learning (DIL) poses a significant challenge in real-world scenarios, as models need to be sequentially trained on diverse domains over time, all the while avoiding catastrophic forgetting. Mitigating representation…

Machine Learning · Computer Science 2024-06-25 Kishaan Jeeveswaran , Elahe Arani , Bahram Zonooz

Accurate uncertainty quantification is crucial for the safe deployment of machine learning models, and prior research has demonstrated improvements in the calibration of modern language models (LMs). We study in-context learning (ICL), a…

Computation and Language · Computer Science 2024-03-29 Hanlin Zhang , Yi-Fan Zhang , Yaodong Yu , Dhruv Madeka , Dean Foster , Eric Xing , Himabindu Lakkaraju , Sham Kakade

Despite significant advances, continual learning models still suffer from catastrophic forgetting when exposed to incrementally available data from non-stationary distributions. Rehearsal approaches alleviate the problem by maintaining and…

Machine Learning · Statistics 2021-03-02 Binh Tang , David S. Matteson

Federated learning (FL) is a burgeoning distributed machine learning framework where a central parameter server (PS) coordinates many local users to train a globally consistent model. Conventional federated learning inevitably relies on a…

Machine Learning · Computer Science 2023-12-13 Zhikun Chen , Daofeng Li , Jinkang Zhu , Sihai Zhang

We introduce Progressive Prompts - a simple and efficient approach for continual learning in language models. Our method allows forward transfer and resists catastrophic forgetting, without relying on data replay or a large number of…

Computation and Language · Computer Science 2023-01-31 Anastasia Razdaibiedina , Yuning Mao , Rui Hou , Madian Khabsa , Mike Lewis , Amjad Almahairi

Pre-trained vision-language models (VLMs) have shown impressive performance on various downstream tasks by utilizing knowledge learned from large data. In general, the performance of VLMs on target tasks can be further improved by prompt…

Computer Vision and Pattern Recognition · Computer Science 2023-09-08 Eulrang Cho , Jooyeon Kim , Hyunwoo J. Kim

A key challenge of continual reinforcement learning (CRL) in dynamic environments is to promptly adapt the RL agent's behavior as the environment changes over its lifetime, while minimizing the catastrophic forgetting of the learned…

Machine Learning · Computer Science 2023-05-25 Tiantian Zhang , Zichuan Lin , Yuxing Wang , Deheng Ye , Qiang Fu , Wei Yang , Xueqian Wang , Bin Liang , Bo Yuan , Xiu Li

Decoupled learning is a branch of model parallelism which parallelizes the training of a network by splitting it depth-wise into multiple modules. Techniques from decoupled learning usually lead to stale gradient effect because of their…

Machine Learning · Computer Science 2020-12-08 Huiping Zhuang , Zhiping Lin , Kar-Ann Toh

Recent advancements in multimodal foundation models (e.g., CLIP) have excelled in zero-shot generalization. Prompt tuning involved in the knowledge transfer from foundation models to downstream tasks has gained significant attention…

Computer Vision and Pattern Recognition · Computer Science 2023-12-07 Xuejing Liu , Wei Tang , Jinghui Lu , Rui Zhao , Zhaojun Guo , Fei Tan

This paper introduces two novel modifications to the Dynamic sAmpling Policy Optimization (DAPO) algorithm [1], approached from a mixed-policy perspective. Standard policy gradient methods can suffer from instability and sample…

Machine Learning · Computer Science 2025-08-20 Hongze Tan , Yuchen Li

Dynamical System (DS) based Learning from Demonstration (LfD) allows learning of reactive motion policies with stability and convergence guarantees from a few trajectories. Yet, current DS learning techniques lack the flexibility to…

Robotics · Computer Science 2023-09-06 Tianyu Li , Nadia Figueroa

Recent advances in multimodal learning has resulted in powerful vision-language models, whose representations are generalizable across a variety of downstream tasks. Recently, their generalization ability has been further extended by…

Computer Vision and Pattern Recognition · Computer Science 2023-12-13 Koustava Goswami , Srikrishna Karanam , Prateksha Udhayanan , K J Joseph , Balaji Vasan Srinivasan

Classical machine learners are designed only to tackle one task without capability of adopting new emerging tasks or classes whereas such capacity is more practical and human-like in the real world. To address this shortcoming, continual…

Machine Learning · Computer Science 2021-12-06 Xuejun Han , Yuhong Guo

Recent rehearsal-free continual learning (CL) methods guided by prompts achieve strong performance on vision tasks with non-stationary data but remain resource-intensive, hindering real-world edge deployment. We introduce resource-efficient…

Machine Learning · Computer Science 2025-12-17 Sungho Jeon , Xinyue Ma , Kwang In Kim , Myeongjae Jeon

Prompt-tuning (PT) for large language models (LLMs) can facilitate the performance on various conventional NLP tasks with significantly fewer trainable parameters. However, our investigation reveals that PT provides limited improvement and…

Computation and Language · Computer Science 2025-04-15 Sinan Fan , Liang Xie , Chen Shen , Ge Teng , Xiaosong Yuan , Xiaofeng Zhang , Chenxi Huang , Wenxiao Wang , Xiaofei He , Jieping Ye

Modern supervised semantic segmentation methods are usually finetuned based on the supervised or self-supervised models pre-trained on ImageNet. Recent work shows that transferring the knowledge from CLIP to semantic segmentation via prompt…

Computer Vision and Pattern Recognition · Computer Science 2023-08-15 Chaohui Yu , Qiang Zhou , Zhibin Wang , Fan Wang