3 | Liyuan Liu https://liyuanlucasliu.github.io/publication_types/3/ 3 Source Themes Academic (https://sourcethemes.com/academic/)en-usCopyright © Liyuan LiuMon, 11 Aug 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/img/avatar.jpg 3 https://liyuanlucasliu.github.io/publication_types/3/ FlashRL: 8Bit Rollouts, Full Power RL https://liyuanlucasliu.github.io/publication/liu-quant-2025/ Mon, 11 Aug 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-quant-2025/ Improving MoE Post-Training with Precise Router Gradients https://liyuanlucasliu.github.io/publication/feng-densemixer-2025/ Tue, 05 Aug 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/feng-densemixer-2025/ Your Efficient RL Framework Secretly Brings You Off-Policy RL Training https://liyuanlucasliu.github.io/publication/feng-off-2025/ Tue, 05 Aug 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/feng-off-2025/ Decoder-Hybrid-Decoder Architecture for Efficient Reasoning with Long Generation https://liyuanlucasliu.github.io/publication/ren-hybrid-2025/ Wed, 09 Jul 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/ren-hybrid-2025/ Training Language Models to Generate Quality Code with Program Analysis Feedback https://liyuanlucasliu.github.io/publication/feng-real-2025/ Wed, 28 May 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/feng-real-2025/ Reinforcement Learning for Reasoning in Large Language Models with One Training Example https://liyuanlucasliu.github.io/publication/wang-onesample-2025/ Sun, 25 May 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/wang-onesample-2025/ Text Generation Beyond Discrete Token Sampling https://liyuanlucasliu.github.io/publication/zhuang-moi-2025/ Tue, 20 May 2025 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhuang-moi-2025/ LoRC: Low-Rank Compression for LLMs KV Cache with a Progressive Compression Strategy https://liyuanlucasliu.github.io/publication/zhang-loca-2024/ Fri, 04 Oct 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhang-loca-2024/ Vector-ICL: In-context Learning with Continuous Vector Representations https://liyuanlucasliu.github.io/publication/zhuang-vectoricl-2025/ Fri, 04 Oct 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhuang-vectoricl-2025/ GRIN: GRadient-INformed MoE https://liyuanlucasliu.github.io/publication/liu-grin-2024/ Wed, 18 Sep 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-grin-2024/ Model Tells Itself Where to Attend: Faithfulness Meets Automatic Attention Steering https://liyuanlucasliu.github.io/publication/zhang-autopasta-2024/ Mon, 16 Sep 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhang-autopasta-2024/ Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone https://liyuanlucasliu.github.io/publication/abdin-phi-2024/ Fri, 30 Aug 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/abdin-phi-2024/ Learning a Decision Tree Algorithm with Transformers https://liyuanlucasliu.github.io/publication/zhuang-autotree-2024/ Tue, 06 Feb 2024 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhuang-autotree-2024/ Sparse Backpropagation for MoE Training https://liyuanlucasliu.github.io/publication/liu-sparsemixer-2023/ Sun, 01 Oct 2023 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-sparsemixer-2023/ Piled: An identify-and-localize framework for few-shot event detection https://liyuanlucasliu.github.io/publication/li-pile-2022/ Tue, 01 Feb 2022 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/li-pile-2022/ UCPhrase: Unsupervised Context-aware Quality Phrase Tagging https://liyuanlucasliu.github.io/publication/gu-ucphrase-2021/ Sun, 01 Aug 2021 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/gu-ucphrase-2021/ Multi-head or Single-head? An Empirical Comparison for Transformer Training https://liyuanlucasliu.github.io/publication/liu-singlehead-2021/ Thu, 17 Jun 2021 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-singlehead-2021/ On the Transformer Growth for Progressive BERT Training https://liyuanlucasliu.github.io/publication/gu-progressive-2020/ Thu, 01 Oct 2020 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/gu-progressive-2020/ Overfitting or Underfitting? Understand Robustness Drop in Adversarial Training https://liyuanlucasliu.github.io/publication/li-apart-2020/ Thu, 01 Oct 2020 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/li-apart-2020/ Very Deep Transformers for Neural Machine Translation https://liyuanlucasliu.github.io/publication/liu-deep-2020/ Sat, 01 Aug 2020 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-deep-2020/ Raw-to-End Name Entity Recognition in Social Media https://liyuanlucasliu.github.io/publication/liu-raw-end-2019/ Thu, 01 Aug 2019 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/liu-raw-end-2019/ Expert Finding in Heterogeneous Bibliographic Networks with Locally-trained Embeddings https://liyuanlucasliu.github.io/publication/gui-expert-2018/ Thu, 01 Mar 2018 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/gui-expert-2018/ Contrast Subgraph Mining from Coherent Cores https://liyuanlucasliu.github.io/publication/shang-contrast-2018/ Thu, 01 Feb 2018 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/shang-contrast-2018/ Wikidata Vandalism Detection - The Loganberry Vandalism Detector at WSDM Cup 2017 https://liyuanlucasliu.github.io/publication/zhu-wikidata-2017/ Fri, 01 Dec 2017 00:00:00 +0000 https://liyuanlucasliu.github.io/publication/zhu-wikidata-2017/