相关论文: Correcting Suppressed Log-Probabilities in Languag…

While reasoning models are increasingly ubiquitous, the effects of reasoning training on a model's internal mechanisms remain poorly understood. In this work, we introduce transcoder adapters, a technique for learning an interpretable…

Adding memory to pretrained language models typically requires architectural changes or weight modification. We present Prometheus Mind, which retrofits memory to a frozen Qwen3-4B using 11 modular adapters (530MB, 7% overhead) -- fully…

We analyze the operation of transformer language adapters, which are small modules trained on top of a frozen language model to adapt its predictions to new target languages. We show that adapted predictions mostly evolve in the source…

Self-interpretation methods prompt language models to describe their own internal states, but remain unreliable due to hyperparameter sensitivity. We show that training lightweight adapters on interpretability artifacts, while keeping the…

Large language models (LMs) are typically adapted to improve performance on new contexts (\eg text prompts that define new tasks or domains) through fine-tuning or prompting. However, there is an accuracy compute tradeoff -- fine-tuning…

Large language models may encounter factual knowledge during pre-training yet fail to reliably use that knowledge after fine-tuning. Despite growing empirical evidence that MLP layers store factual associations and fine-tuning affects…

Adapter-tuning is a paradigm that transfers a pretrained language model to downstream tasks by adding and tuning a small number of new parameters. Previously proposed adapter architectures are all feed-forward neural networks. In this…

Large language models fine-tuned via a two-stage pipeline (domain adaptation followed by instruction alignment) can exhibit non-trivial interference after adapter merging, including the re-emergence of explicit reasoning traces under strict…

While random demonstration labels barely hurt in-context learning (Min et al., 2022), we show that homogeneous labels--even semantically valid ones--collapse accuracy to <=12% across six models (Pythia, Llama, Qwen; 0.8B--8B) and four…

Mathematical reasoning models are widely deployed in education, automated tutoring, and decision support systems despite exhibiting fundamental computational instabilities. We demonstrate that state-of-the-art models (Qwen2.5-Math-7B)…

Why do language models trained on contradictory data prefer correct answers? In controlled experiments with small transformers (3.5M--86M parameters), we show that this preference tracks the compressibility structure of errors rather than…

Adapter-based Federated Large Language Models (FedLLMs) are widely adopted to reduce the computational, storage, and communication overhead of full-parameter fine-tuning for web-scale applications while preserving user privacy. By freezing…

Most commonly used language models (LMs) are instruction-tuned and aligned using a combination of fine-tuning and reinforcement learning, causing them to refuse users requests deemed harmful by the model. However, jailbreak prompts can…

Looped Transformers offer a promising alternative to purely feed-forward computation by iteratively refining latent representations, improving language modeling and reasoning. Yet recurrent architectures remain unstable to train, costly to…

We present a language model that combines a large parametric neural network (i.e., a transformer) with a non-parametric episodic memory component in an integrated architecture. Our model uses extended short-term context by caching local…

Recently, the pre-trained Transformer models have received a rising interest in the field of speech processing thanks to their great success in various downstream tasks. However, most fine-tuning approaches update all the parameters of the…

With excellent generalization ability, self-supervised speech models have shown impressive performance on various downstream speech tasks in the pre-training and fine-tuning paradigm. However, as the growing size of pre-trained models,…

State-of-the-art pretrained NLP models contain a hundred million to trillion parameters. Adapters provide a parameter-efficient alternative for the full finetuning in which we can only finetune lightweight neural network layers on top of…

Sequential fine-tuning of pretrained language encoders often overwrites previously acquired capabilities, but the forgetting behavior of parameter-efficient updates remains under-characterized. We present a controlled empirical study of…

Decoder-only language models are stateless: hidden representations are discarded after every forward pass and nothing persists across sessions. Jeong (2026a) showed that trained memory adapters give a frozen encoder-decoder backbone…