Matryoshka: Learning to Drive Black-Box LLMs with LLMs
- URL: http://arxiv.org/abs/2410.20749v1
- Date: Mon, 28 Oct 2024 05:28:51 GMT
- Title: Matryoshka: Learning to Drive Black-Box LLMs with LLMs
- Authors: Changhao Li, Yuchen Zhuang, Rushi Qiang, Haotian Sun, Hanjun Dai, Chao Zhang, Bo Dai,
- Abstract summary: Matryoshika is a lightweight white-box large language models controller.
It guides a large-scale black-box LLM generator by decomposing complex tasks into a series of intermediate outputs.
- Score: 31.501244808646
- License:
- Abstract: Despite the impressive generative abilities of black-box large language models (LLMs), their inherent opacity hinders further advancements in capabilities such as reasoning, planning, and personalization. Existing works aim to enhance LLM capabilities via domain-specific adaptation or in-context learning, which require additional training on accessible model parameters, an infeasible option for black-box LLMs. To address this challenge, we introduce Matryoshika, a lightweight white-box LLM controller that guides a large-scale black-box LLM generator by decomposing complex tasks into a series of intermediate outputs. Specifically, we consider the black-box LLM as an environment, with Matryoshika serving as a policy to provide intermediate guidance through prompts for driving the black-box LLM. Matryoshika is trained to pivot the outputs of the black-box LLM aligning with preferences during iterative interaction, which enables controllable multi-turn generation and self-improvement in optimizing intermediate guidance. Empirical evaluations on three diverse tasks demonstrate that Matryoshika effectively enhances the capabilities of black-box LLMs in complex, long-horizon tasks, including reasoning, planning, and personalization. By leveraging this pioneering controller-generator framework to mitigate dependence on model parameters, Matryoshika provides a transparent and practical solution for improving black-box LLMs through controllable multi-turn generation using white-box LLMs.
Related papers
- LLM-Lasso: A Robust Framework for Domain-Informed Feature Selection and Regularization [59.75242204923353]
We introduce LLM-Lasso, a framework that leverages large language models (LLMs) to guide feature selection in Lasso regression.
LLMs generate penalty factors for each feature, which are converted into weights for the Lasso penalty using a simple, tunable model.
Features identified as more relevant by the LLM receive lower penalties, increasing their likelihood of being retained in the final model.
arXiv Detail & Related papers (2025-02-15T02:55:22Z) - LLMs can see and hear without any training [63.964888082106974]
MILS is a simple, training-free approach to imbue multimodal capabilities into your favorite LLM.
We establish a new state-of-the-art on emergent zero-shot image, video and audio captioning.
Being a gradient-free optimization approach, MILS can invert multimodal embeddings into text.
arXiv Detail & Related papers (2025-01-30T02:16:35Z) - OLA-VLM: Elevating Visual Perception in Multimodal LLMs with Auxiliary Embedding Distillation [95.78870389271832]
The standard practice for developing contemporary MLLMs is to feed features from vision encoder(s) into the LLM and train with natural language supervision.
We propose OLA-VLM, the first approach distilling knowledge into the LLM's hidden representations from a set of target visual representations.
We show that OLA-VLM boosts performance by an average margin of up to 2.5% on various benchmarks, with a notable improvement of 8.7% on the Depth task in CV-Bench.
arXiv Detail & Related papers (2024-12-12T18:55:18Z) - AIM: Adaptive Inference of Multi-Modal LLMs via Token Merging and Pruning [19.68349294206012]
We propose a training-free adaptive inference method for multi-modal LLMs.
With a minimalist design, our method can be applied to both video and image LLMs.
Under a similar computational cost, our method outperforms the state-of-the-art methods in long video understanding.
arXiv Detail & Related papers (2024-12-04T11:47:57Z) - LLM-wrapper: Black-Box Semantic-Aware Adaptation of Vision-Language Models for Referring Expression Comprehension [45.856469849910496]
We propose a method for 'black-box' adaptation of Vision Language Models (VLMs) for the Referring Expression (REC) task using Large Language Models (LLMs)
LLMs capitalizes on the reasoning abilities of LLMs, improved with a light fine-tuning, to select the most relevant bounding box matching the referring expression.
Our approach offers several advantages: it enables the adaptation of closed-source models without needing access to their internal workings.
arXiv Detail & Related papers (2024-09-18T12:32:25Z) - SEA: Supervised Embedding Alignment for Token-Level Visual-Textual Integration in MLLMs [40.74693126923826]
Multimodal Large Language Models (MLLMs) have recently demonstrated remarkable perceptual and reasoning abilities.
Training adapters with image-level supervision often results in significant misalignment.
We introduce Supervised Embedding Alignment (SEA), a token-level alignment method that leverages vision-language pre-trained models.
arXiv Detail & Related papers (2024-08-21T17:58:02Z) - Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing [56.75702900542643]
We introduce AlphaLLM for the self-improvements of Large Language Models.
It integrates Monte Carlo Tree Search (MCTS) with LLMs to establish a self-improving loop.
Our experimental results show that AlphaLLM significantly enhances the performance of LLMs without additional annotations.
arXiv Detail & Related papers (2024-04-18T15:21:34Z) - From Image to Video, what do we need in multimodal LLMs? [19.85928004619801]
Multimodal Large Language Models (MLLMs) have demonstrated profound capabilities in understanding multimodal information.
We propose RED-VILLM, a Resource-Efficient Development pipeline for Video LLMs from Image LLMs.
Our approach highlights the potential for a more cost-effective and scalable advancement in multimodal models.
arXiv Detail & Related papers (2024-04-18T02:43:37Z) - Towards Vision Enhancing LLMs: Empowering Multimodal Knowledge Storage
and Sharing in LLMs [72.49064988035126]
We propose an approach called MKS2, aimed at enhancing multimodal large language models (MLLMs)
Specifically, we introduce the Modular Visual Memory, a component integrated into the internal blocks of LLMs, designed to store open-world visual information efficiently.
Our experiments demonstrate that MKS2 substantially augments the reasoning capabilities of LLMs in contexts necessitating physical or commonsense knowledge.
arXiv Detail & Related papers (2023-11-27T12:29:20Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.