Entity · paper

Disentangling Curriculum Learning in NLP: Towards a Unifying Taxonomy

paperactiveprovisionaldisentangling-curriculum-learning-in-nlp-towards-a-unifying-taxonomy-fb1d2245·1 events·first seen 5d ago

Aliases: Disentangling Curriculum Learning in NLP: Towards a Unifying Taxonomy

More like this (12)

From Isolated Tasks to Structured Capabilities: A Multilayer Taxonomy for Large Language Models Beyond Surface Forms: A Comprehensive, Mechanism-Oriented Taxonomy of Indirect Linguistic Encoding for LLM-Based Coded Language Detection Detecting Knowledge Gaps from Conversational AI Interactions Using Curriculum Prerequisite Graphs curriculum learning TILDE: TILt-based Distributional Erasure for Concept Unlearning Same Lesson, Different Story: Cross-Lingual Reconstruction of Cultural Narratives in Large Language Models Disentangled Representation Learning Meta-Learning for Compositionality How Does Research Evolve? Tracing Cross-Domain Trajectories in NLP, ML, and CV with Claim-Grounded Typed Citations Natural Language Processing Arithmetic Pedagogy for Language Models Automated grading of Linux/bash examinations using large language models: a four-level cognitive taxonomy approach

Recent events (1)

4arXiv · cs.CL·5d ago·source ↗

Taxonomy paper proposes unified framework for curriculum learning in NLP

A new arXiv preprint proposes a fine-grained taxonomy for curriculum learning (CL) in NLP, separating difficulty evaluation from training scheduling to enable systematic comparison of CL strategies. The authors identify a 'systematic incomparability problem' in prior work, where different notions of difficulty and scheduling are conflated under the same CL label. The paper provides the first formal treatment of CL schedulers in terms of expected training contribution, introducing retention regimes and monotonicity properties. The work is primarily a conceptual and analytical contribution aimed at improving rigor in CL research rather than reporting new empirical results.

Alignment and RLHF Disentangling Curriculum Learning in NLP: Towards a Unifying Taxonomy