PyTorchKR Papers

최신 논문

총 2편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2606.11854v2cs.LG

ART를 통한 멀티모달 LLM 미세조정: 아트 기반 강화 학습

Fine-tuning Multi-modal LLMs with ART: Art-based Reinforcement Training

Adapting Large Language Models (LLMs) to specific tasks usually involves fine-tuning: Low-Rank Adaptation (LoRA) edits the weights, while soft prompting prepends trainable virtual tokens. Both change how an already-optimized model is served, so serving many task-specific adapters at high throughput requires adapter-aware infrastructure. We propose Art-Based Reinforcement Training (ART), which optimizes a single auxiliary input image to specialize a frozen Multimodal Large Language Model (MLLM) for a specific task. At inference time, ART simply includes this image in each standard multimodal prompt, requiring no changes to the serving engine. We compare ART with LoRA and soft prompting across five reasoning benchmarks for four small MLLMs from three architectures. While different methods perform best for different tasks, ART ranks first or second in 10 of 20 benchmark-MLLM pairs, with its best results on the two math word problem benchmarks (GSM8k and SVAMP). It is significantly ahead of the next best adapter in 2 cases, within the noise of the next best adapter in 6, and significantly above the text-only baseline in 18. ART therefore reaches accuracy close to LoRA and soft prompts while serving faster than LoRA. In the vLLM serving stack, ART reduces throughput by 6% to 14% relative to text-only requests, compared to LoRA's 13% to 93% slowdown. Beyond efficient serving, ART artifacts are plain images, making them easy to store, distribute, and share. This approach keep task-specific adaptation outside the model weights, so the same frozen model can support different tasks without separate model variants.

arXiv:2607.06145v1cs.CL

프롬프팅 복잡도: LLM에서 텍스트와 행동을 위한 최단 프롬프트

Prompting Complexity: Shortest Prompts for Texts and Behaviors in LLMs

고정된 인스트럭션 튜닝 대규모 언어 모델(LLM)에서 목표 텍스트를 생성하는 가장 짧은 그럴듯한 프롬프트의 길이를 측정하는 프롬프팅 복잡도(prompting complexity)를 자원 제한 콜모고로프 복잡도(Kolmogorov complexity)의 언어 모델 상대적 유사체로 형식화하였다. 근사 출력을 위한 소프트 프롬프팅 복잡도, 프롬프팅 거리, 행동적 프롬프팅 복잡도로 개념을 확장하며, 어떤 텍스트와 행동이 짧은 프롬프트로 접근 가능한지를 실증적으로 연구하기 위한 연구 의제를 제안한다.