arXiv AI September 3, 2026 By Zhao Ji, Wenqing Chen, Zhixuan Chu, Jianxing Yu, Jingping Liu, Shanhe Zhao, Zibin Zheng

SALA: Semantic-Aware Logical Alignment for Complex Reasoning in In-Context Learning

The paper introduces SALA, a Semantic‑Aware Logical Alignment framework designed to improve demonstration selection for complex reasoning in in‑context learning. SALA learns task‑specific reasoning operations, embeds them into a continuous semantic space, and applies dynamic time warping to flexibly align reasoning sequences, offering soft matching and interpretability. Experiments on four reasoning benchmarks with three large language models show that SALA outperforms existing methods, and analysis highlights the importance of operation induction and logical semantic alignment.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI

Jun 2

Many-Shot CoT-ICL: Making In-Context Learning Truly Learn

arXiv:2605. 13511v3 Announce Type: replace-cross Abstract: While many-shot ICL achieves remarkable performance, prior studies of its scaling behavior have mainly focused on non-reasoning tasks.

By Tsz Ting Chung, Lemao Liu, Mo Yu, Dit-Yan Yeung

llms

arXiv AI

Aug 18

QA-Merging: Query-Adaptive Reasoning via Layer Selective Model Merging

arXiv:2601. 03506v2 Announce Type: replace-cross Abstract: Recent large reasoning models (LRMs) have achieved strong performance on complex reasoning tasks by generating a long chain-of-thought (Long-CoT).

By Zhaofeng Zhong, Wei Yuan, Tong Chen, Liang Qu, Xiangyu Zhao, Quoc Viet Hung Nguyen, Hongzhi Yin

llms benchmarks safety

arXiv AI

Oct 21

CHIMERA: Compact Synthetic Data for Generalizable LLM Reasoning

arXiv:2603.00889v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have exhibited remarkable reasoning capabilities, largely enabled by supervised fine-tuning (SFT)- and reinforce...

By Xinyu Zhu, Yihao Feng, Yanchao Sun, Xianzhi Du, Pingzhi Li, Olli Saarikivi, Yun Zhu, Yu Meng

llms reinforcement-learning fine-tuning benchmarks

arXiv AI

Sep 11

Structural Process Supervision for Latent Chain-of-Thought Reasoning

The paper introduces Prototype-Mediated Process Supervision (PMPS), a method that uses learnable reasoning prototypes to provide structural supervision for latent chain-of-thought embeddings, addressing representation collapse and uneven information distribution. PMPS aligns latent and explicit CoT embeddings in a shared prototype space via many-to-many soft assignment and employs a Progressive Sequential Alignment module to guide training from positional priors to adaptive matching. Experiments show PMPS reduces output token length to under 50% of explicit CoT on GSM8K-Aug and improves accuracy by 2.08% over SIM-CoT, even surpassing CoT-SFT on GPT-2 and achieving the highest accuracy among latent reasoning methods on larger models and harder tasks.

By Yiqi Li, Xu Chen, Chen Ju, Jiangchao Yao, Zhaoyang Li, Jinsong Lan, Xiaoyong Zhu, Bo Zheng, Yu Wang

llms rag fine-tuning safety

arXiv AI

Jun 2

PATRA: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering

arXiv:2602. 23161v4 Announce Type: replace Abstract: Time series reasoning demands both the perception of complex dynamics and logical depth.

By Junkai Lu, Peng Chen, Xingjian Wu, Yang Shu, Chenjuan Guo, Christian S. Jensen, Bin Yang

llms nlp multimodal safety

arXiv AI

Oct 21

Attention-Aligned Reasoning for Large Language Models

arXiv:2510.03223v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) tend to generate a long reasoning chain when solving complex tasks. However, as the reasoning chain extends, cri...

By Hongxiang Zhang, Yuan Tian, Tianyi Zhang

llms benchmarks safety

One email a morning, machine-written

Related stories

Many-Shot CoT-ICL: Making In-Context Learning Truly Learn

QA-Merging: Query-Adaptive Reasoning via Layer Selective Model Merging

CHIMERA: Compact Synthetic Data for Generalizable LLM Reasoning

Structural Process Supervision for Latent Chain-of-Thought Reasoning

PATRA: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering

Attention-Aligned Reasoning for Large Language Models