--- title: "Yang Liu — papers, page 4 — Elamur Research" description: "Yang Liu, Beijing University of Posts and Telecommunications: 81 papers on Elamur Research (2006–2026), h-index 1 — affiliations, topics…" canonical: "https://research.elamur.ai/free/en/authors/yang-liu/page/4/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/yang-liu/page/4/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / [Yang Liu](https://md.research.elamur.ai/free/en/authors/yang-liu/) / page 4 # Yang Liu · page 4 81 papers in our library · 2006–2026 · page 4 of 4 [The profile, affiliations and co-authors →](https://md.research.elamur.ai/free/en/authors/yang-liu/) ## Papers The most cited first (OpenAlex and our library). [← prev](https://md.research.elamur.ai/free/en/authors/yang-liu/page/3/) [1](https://md.research.elamur.ai/free/en/authors/yang-liu/) [2](https://md.research.elamur.ai/free/en/authors/yang-liu/page/2/) [3](https://md.research.elamur.ai/free/en/authors/yang-liu/page/3/) **4** 76. **[Plug-and-Play Data Module for Code RL: Adaptive Ambiguity Replay](https://md.research.elamur.ai/free/en/papers/zhang-2026-plug-play-data-module-code/)** *(2026)* Reinforcement learning helps LLMs generate and edit code, but good training data is scarce. Experience replay can stretch limited data. However, current LLM replay methods such as RePO and RLEP choose what to replay using static, single-epoch signals like reward or variance, and they can feed the model stale… 77. **[Rethinking Data Mixing from the Perspective of Large Language Models](https://md.research.elamur.ai/free/en/papers/xu-2026-rethinking-data-mixing-perspective-large/)** *(2026)* The paper asks how training data from different domains should be weighted when pretraining LLMs. It also asks whether human-assigned domain labels (e.g. C4, Wikipedia, ArXiv, Book) match what the model itself treats as distinct. The authors observe that per-sample gradients form clearly separated clusters by source… 78. **[Revisiting a Pain in the Neck: A Semantic Reasoning Benchmark for Language Models](https://md.research.elamur.ai/free/en/papers/liu-2026-revisiting-pain-neck-semantic-reasoning/)** *(2026)* SemanticQA is an English benchmark that tests how well language models handle semantic phrases, also called multiword expressions (MWE). 79. **[The Art of (Mis)alignment: How Fine-Tuning Methods Effectively Misalign and Realign LLMs in Post-Training](https://md.research.elamur.ai/free/en/papers/zhang-2026-art-mis-alignment-how-fine/)** *(2026)* Safety alignment tools cut both ways. The fine-tuning that teaches an LLM to refuse harmful requests can also be used to remove those refusals, and the compromised models can then be shared on open hubs. The authors frame this as an attacker–defender game. They ask which fine-tuning methods most effectively misalign… 80. **[Trustworthy and Explainable Causal Representation Learning in Transformers](https://md.research.elamur.ai/free/en/papers/liu-2026-trustworthy-explainable-causal-representation-learning/)** *(2026)* Explanations for Transformers are often built by learning an importance mask over input features and applying it to the input representation. 81. **[Understanding the Prompt Sensitivity](https://md.research.elamur.ai/free/en/papers/liu-2026-understanding-prompt-sensitivity/)** *(2026)* Large language models often give different answers to prompts that mean the same thing, yet the reason for this has been unclear. [← prev](https://md.research.elamur.ai/free/en/authors/yang-liu/page/3/) [1](https://md.research.elamur.ai/free/en/authors/yang-liu/) [2](https://md.research.elamur.ai/free/en/authors/yang-liu/page/2/) [3](https://md.research.elamur.ai/free/en/authors/yang-liu/page/3/) **4**