13 papers
Efficient Reasoning Training Does Not Always Harm CoT Faithfulness and Monitorability
Samuel Lewis-Lim, Xingwei Tan, Mario Sanger +2
Chain-of-thought (CoT) reasoning allows humans to inspect how large language models reach their answers, and oversee model behaviour. This reasoning comes at an increased inference…
Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior
Atsuki Yamaguchi, Tatsuro Inaba, Joel Niklaus +3
Pre-pretraining (PPT) on synthetic non-natural language data improves token efficiency during language model pre-training (PT). Prior work attributes this gain to a grammatical pri…
BusMA: A Bus Communication Substrate for Multi-Agent Systems
Yanwen Peng, Delvin Ce Zhang, Xi Wang +1
Multi-Agent (MA) systems are effective at solving complex tasks that demand planning, tool use, and the synthesis of evidence from multiple sources. Existing systems typically adop…
StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems
Yanwen Peng, Delvin Ce Zhang, Xi Wang +1
Large language model based multi-agent systems usually communicate in text, i.e., using discrete tokens. However, text introduces a discrete bottleneck. Converting the sender's con…
Logic Before Language: Pre-pretraining on Formal Derivations Fosters Skill Acquisition and Compressibility
Jo-Ku Cheng, Nikolaos Aletras, Marco Valentino
Pre-pretraining language models (LMs) on symbolic data can accelerate and improve natural language acquisition. However, existing pre-pretraining tasks, such as Dyck and procedural…
On Effectiveness and Efficiency of Agentic Tool-calling and RL Training
Tong Liu, Cheng Qian, Matej Cief +4
Tool-calling is a central component of modern large language model (LLM) agents, equipping them with skills beyond their parametric knowledge. This paper studies tool-calling along…
Compliance vs. Sensibility: On the Reasoning Controllability in Large Language Models
Xingwei Tan, Marco Valentino, Mahmud Elahi Akhter +3
Large Language Models (LLMs) acquire reasoning capabilities through shared inference patterns in pre-training data, which are further elicited via Chain-of-Thought (CoT). However,…
Reasoning Dynamics and the Limits of Monitoring Modality Reliance in Vision-Language Models
Danae Sánchez Villegas, Samuel Lewis-Lim, Nikolaos Aletras +1
Recent advances in vision language models (VLMs) offer reasoning capabilities, yet how these unfold and integrate visual and textual information remains unclear. We analyze reasoni…
Fundamental Reasoning Paradigms Induce Out-of-Domain Generalization in Language Models
Mingzi Cao, Xingwei Tan, Mahmud Elahi Akhter +4
Deduction, induction, and abduction are fundamental reasoning paradigms, core for human logical thinking. Although improving Large Language Model (LLM) reasoning has attracted sign…
An Empirical Study on Preference Tuning Generalization and Diversity Under Domain Shift
Constantinos Karouzos, Xingwei Tan, Nikolaos Aletras
Preference tuning aligns base language models to human judgments of quality, helpfulness, or safety by optimizing over explicit preference signals rather than likelihood alone. Pri…
Fine-Tuning on Noisy Instructions: Effects on Generalization and Performance
Ahmed Alajrami, Xingwei Tan, Nikolaos Aletras
Instruction-tuning plays a vital role in enhancing the task-solving abilities of large language models (LLMs), improving their usability in generating helpful responses on various…
Can Confidence Estimates Decide When Chain-of-Thought Is Necessary for LLMs?
Samuel Lewis-Lim, Xingwei Tan, Zhixue Zhao +1
Chain-of-thought (CoT) prompting is a common technique for improving the reasoning abilities of large language models (LLMs). However, extended reasoning is often unnecessary and s…
PATCH: Mitigating PII Leakage in Language Models with Privacy-Aware Targeted Circuit PatcHing
Anthony Hughes, Vasisht Duddu, N. Asokan +2
Language models (LMs) may memorize personally identifiable information (PII) from training data, enabling adversaries to extract it during inference. Existing defense mechanisms su…