Related papers: Crosslingual On-Policy Self-Distillation for Multi…
Large language models (LLMs) have garnered increasing attention owing to their powerful logical reasoning capabilities. Generally, larger LLMs (L-LLMs) that require paid interfaces exhibit significantly superior performance compared to…
Researchers working on low-resource languages face persistent challenges due to limited data availability and restricted access to computational resources. Although most large language models (LLMs) are predominantly trained in…
Recently, multi-modal content generation has attracted lots of attention from researchers by investigating the utilization of visual instruction tuning based on large language models (LLMs). To enhance the performance and generalization…
Large reasoning models achieve strong performance on complex tasks through long chain-of-thought (CoT) trajectories, but directly transferring such reasoning processes to smaller models remains challenging. A key difficulty is that not all…
In this work, we study whether multilingual language models (MultiLMs) can transfer logical reasoning abilities to other languages when they are fine-tuned for reasoning in a different language. We evaluate the cross-lingual reasoning…
Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, their enormous parameter size and extremely high requirements for compute power pose challenges for…
Knowledge Distillation (KD) compresses computationally expensive pre-trained language models (PLMs) by transferring their knowledge to smaller models, allowing their use in resource-constrained or real-time settings. However, most smaller…
Large language models (LLMs) are known to effectively perform tasks by simply observing few exemplars. However, in low-resource languages, obtaining such hand-picked exemplars can still be challenging, where unsupervised techniques may be…
Although all-in-one-model multilingual neural machine translation (multilingual NMT) has achieved remarkable progress, the convergence inconsistency in the joint training is ignored, i.e., different language pairs reaching convergence in…
On-policy distillation (OPD) trains a student on its own trajectories under token-level teacher supervision, but existing methods are capped by a single-teacher capability ceiling: when the teacher errs, the student inherits the error. OPD…
While large language models (LLMs) exhibit strong multilingual abilities, their reliance on English as latent representations creates a translation barrier, where reasoning implicitly depends on internal translation into English. When this…
Improving the code generation capabilities of large language models (LLMs) typically relies on supervised fine-tuning or preference optimization, both of which require costly external resources such as powerful teacher models or reliable…
Scaling multilingual representation learning beyond the hundred most frequent languages is challenging, in particular to cover the long tail of low-resource languages. A promising approach has been to train one-for-all multilingual models…
Large language models trained with reinforcement learning (RL) for mathematical reasoning face a fundamental challenge: on problems the model cannot solve at all - "cliff" prompts - the RL gradient vanishes entirely, preventing any learning…
Reinforcement learning from verifiable rewards (RLVR) suffers from sparse outcome signals, creating severe exploration bottlenecks on complex reasoning tasks. Recent on-policy self-distillation methods attempt to address this by utilizing…
In this paper, we investigate how model distillation impacts the development of reasoning features in large language models (LLMs). To explore this, we train a crosscoder on Qwen-series models and their fine-tuned variants. Our results…
Advanced reasoning typically requires Chain-of-Thought prompting, which is accurate but incurs prohibitive latency and substantial test-time inference costs. The standard alternative, fine-tuning smaller models, often sacrifices…
Self-supervised learning (SSL) has made remarkable progress in visual representation learning. Some studies combine SSL with knowledge distillation (SSL-KD) to boost the representation learning performance of small models. In this study, we…
Reinforcement learning has emerged as a powerful tool for improving diffusion-based text-to-image models, but existing methods are largely limited to single-task optimization. Extending RL to multiple tasks is challenging: joint…
Knowledge Distillation (KD) is a fundamental technique for compressing large language models (LLMs) into compact, efficient student models. However, existing white-box KD methods mainly focus on balancing ground truth and student-generated…