Related papers: Transcoder Adapters for Reasoning-Model Diffing
Complex video reasoning, actually, relies excessively on fine-grained perception rather than on expert (e.g., Ph.D, Science)-level reasoning. Through extensive empirical observation, we have recognized the critical impact of perception. In…
The paradigm shift in large language models (LLMs) from instinctive responses to chain-of-thought (CoT) reasoning has fueled two prevailing assumptions: (1) reasoning capabilities only emerge in sufficiently large models, and (2) such…
Fine-tuning of self-supervised models is a powerful transfer learning method in a variety of fields, including speech processing, since it can utilize generic feature representations obtained from large amounts of unlabeled data.…
Cross-lingual speech adaptation aims to solve the problem of leveraging multiple rich-resource languages to build models for a low-resource target language. Since the low-resource language has limited training data, speech recognition…
Transformer language models such as GPT-2 are difficult to quantize because of outliers in activations leading to a large quantization error. To adapt to the error, one must use quantization-aware training, which entails a fine-tuning…
Reasoning is an integral part of many tasks performed by language models (LMs). However, the effects of scaling model sizes and data on reasoning abilities at pretraining time remain understudied. To rigorously investigate this problem, we…
The mechanisms by which reasoning training reshapes LLMs' internal computations remain unclear. We study lightweight steering vectors inserted into the base model's residual stream and trained with a reinforcement-learning objective. These…
Large language models (LLMs) have shown remarkable reasoning capabilities, yet aligning such abilities to small language models (SLMs) remains a challenge due to distributional mismatches and limited model capacity. Existing reasoning…
There is a substantial body of literature examining the mathematical reasoning capabilities of large language models (LLMs), particularly their performance on precise arithmetic operations in autoregressive architectures. However, their…
Recent psycholinguistic research has compared human reading times to surprisal estimates from language models to study the factors shaping human sentence processing difficulty. Previous studies have shown a strong fit between surprisal…
Do reasoning models have "Aha!" moments? Prior work suggests that models like DeepSeek-R1-Zero undergo sudden mid-trace realizations that lead to accurate outputs, implying an intrinsic capacity for self-correction. Yet, it remains unclear…
Can language models (LMs) learn to faithfully describe their internal computations? Are they better able to describe themselves than other models? We study the extent to which LMs' privileged access to their own internals can be leveraged…
Long-sequence processing is a critical capability for modern large language models. However, the self-attention mechanism in the standard Transformer architecture faces severe computational and memory bottlenecks when processing long…
Multilayer perceptrons (MLPs) are an integral part of large language models, yet their dense representations render them difficult to understand, edit, and steer. Recent methods learn interpretable approximations via neuron-level sparsity,…
Reasoning language models, which generate long chains of thought, dramatically outperform non-reasoning language models on abstract problems. However, the internal model mechanisms that allow this superior performance remain poorly…
Fine-tuning pre-trained transformers is a powerful technique for enhancing the performance of base models on specific tasks. From early applications in models like BERT to fine-tuning Large Language Models (LLMs), this approach has been…
Simultaneous machine translation aims at solving the task of real-time translation by starting to translate before consuming the full input, which poses challenges in terms of balancing quality and latency of the translation. The wait-$k$…
Multilingual machine translation suffers from negative interference across languages. A common solution is to relax parameter sharing with language-specific modules like adapters. However, adapters of related languages are unable to…
The capabilities of Large Language Models (LLMs) in low-resource languages lag far behind those in English, making their universal accessibility a significant challenge. To alleviate this, we present $\textit{Franken-Adapter}$, a modular…
Large Language Models have shown impressive capabilities in coding tasks like code generation and code completion, as they have been trained on a large amount of code data. Also, since one of the core pretraining objectives is Next Token…