Profiles2022-2026
Yequan Wang
Beijing Academy of Artificial Intelligence
16Papers
3Last author
28Unique co-authors
1First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers16
- VisualPrompter: Semantic-Aware Prompt Optimization with Visual Feedback for Text-to-Image SynthesisAuthor 4 of 5ICLR 2026 · 7 citations
- Break Through the Compression Bottleneck: From Theory to PracticeAuthor 3 of 5ACL Findings 2026
- Theory-optimal Quantization Based on FlatnessAuthor 5 of 8ACL 2026
- Position-Aware Depth Decay Decoding (D³): Boosting Large Language Model Inference EfficiencyAuthor 6 of 6ACL Findings 2025 · 4 citations
- Capability Localization: Capabilities Can be Localized rather than Individual KnowledgeAuthor 3 of 5ICLR 2025
- SpikeLM: Towards General Spike-Driven Language Modeling via Elastic Bi-Spiking MechanismsAuthor 7 of 9ICML 2024 · 34 citations
- Masked Structural Growth for 2x Faster Language Model Pre-trainingAuthor 4 of 4ICLR 2024 · 30 citations
- Multimodal Reasoning with Multimodal Knowledge GraphAuthor 2 of 4ACL 2024 · 29 citations
- Reasons and Solutions for the Decline in Model Performance after EditingAuthor 3 of 4NeurIPS 2024 · 13 citations
- BiPFT: Binary Pre-trained Foundation Transformer with Low-Rank Estimation of Binarization Residual PolynomialsAuthor 5 of 7AAAI 2024 · 5 citations
- Discerning and Resolving Knowledge Conflicts through Adaptive Decoding with Contextual Information-Entropy ConstraintAuthor 3 of 6ACL Findings 2024 · 4 citations
- Improving Zero-shot LLM Re-Ranker with Risk MinimizationAuthor 3 of 5EMNLP 2024 · 3 citations
- Commonsense Knowledge Editing Based on Free-Text in LLMsAuthor 2 of 4EMNLP 2024
- Rethinking Document-Level Relation Extraction: A Reality CheckAuthor 2 of 4ACL Findings 2023 · 14 citations
- Knowledgeable Parameter Efficient Tuning Network for Commonsense Question AnsweringAuthor 5 of 5ACL 2023 · 4 citations
- CORT: A New Baseline for Comparative Opinion Classification by Dual PromptsAuthor 1 of 4EMNLP Findings 2022 · 4 citations
