conftrace_

pre-training paradigm

5 papers

Explore in graph

Co-occurring keywords · papers mentioning both

sentiment element (1) block-local transformer (1) test-time computation (1) long-input summarization (1) sentiment quadruple (1) graph pre-training (1) left-to-right generation (1) block-local attention (1) tokenizer design (1) post-training strategies (1)

Papers

Compressing then Matching: An Efficient Pre-training Paradigm for Multimodal Embedding ACL 2026 Do LLM Modules Generalize? A Study on Motion Generation for Autonomous Driving CoRL 2025 Exploring Graph Pre-training for Aspect-based Sentiment Analysis EMNLP 2023 Investigating Efficiently Extending Transformers for Long Input Summarization EMNLP 2023 Meet in the Middle: A New Pre-training Paradigm NeurIPS 2023