Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2503.16334 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915365074763776 |
|---|---|
| author | Shen, Ying Huang, Lifu |
| author_facet | Shen, Ying Huang, Lifu |
| contents | Recent findings reveal that much of the knowledge in a Transformer-based Large Language Model (LLM) is encoded in its feed-forward (FFN) layers, where each FNN layer can be interpreted as the summation of sub-updates, each corresponding to a weighted column vector from the FFN's value parameter matrix that often encodes human-interpretable concepts. In light of this, we hypothesize that model performance and behaviors can be further enhanced and controlled by modulating the contributions of these sub-updates based on their relevance to the input or target output style, and propose LLMBRACES, a novel and efficient method that computes relevance scores associated with value vectors in FFN layers and leverages these scores to dynamically adjust the contribution of sub-updates. By optimizing sub-update contributions, LLMBRACES refines the prediction process, leading to more accurate and reliable outputs, much like a 'brace' providing support and stability. Moreover, LLMBRACES can be extended to support conditional control over generation characteristics, such as sentiment, thereby offering fine-grained steering of LLM outputs. Extensive experiments on various LLMs-including Qwen2.5-1.5B, Llama2-7B, and Llama3-8B-demonstrate that LLMBRACES outperforms baseline approaches in both fine-tuning and zero-shot settings while requiring significantly fewer tunable parameters, up to 75% fewer compared to LoRA. Furthermore, LLMBRACES excels in sentiment-controlled generation and toxicity reduction, highlighting its potential for flexible, controlled text generation across applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_16334 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LLM Braces: Straightening Out LLM Predictions with Relevant Sub-Updates Shen, Ying Huang, Lifu Computation and Language Recent findings reveal that much of the knowledge in a Transformer-based Large Language Model (LLM) is encoded in its feed-forward (FFN) layers, where each FNN layer can be interpreted as the summation of sub-updates, each corresponding to a weighted column vector from the FFN's value parameter matrix that often encodes human-interpretable concepts. In light of this, we hypothesize that model performance and behaviors can be further enhanced and controlled by modulating the contributions of these sub-updates based on their relevance to the input or target output style, and propose LLMBRACES, a novel and efficient method that computes relevance scores associated with value vectors in FFN layers and leverages these scores to dynamically adjust the contribution of sub-updates. By optimizing sub-update contributions, LLMBRACES refines the prediction process, leading to more accurate and reliable outputs, much like a 'brace' providing support and stability. Moreover, LLMBRACES can be extended to support conditional control over generation characteristics, such as sentiment, thereby offering fine-grained steering of LLM outputs. Extensive experiments on various LLMs-including Qwen2.5-1.5B, Llama2-7B, and Llama3-8B-demonstrate that LLMBRACES outperforms baseline approaches in both fine-tuning and zero-shot settings while requiring significantly fewer tunable parameters, up to 75% fewer compared to LoRA. Furthermore, LLMBRACES excels in sentiment-controlled generation and toxicity reduction, highlighting its potential for flexible, controlled text generation across applications. |
| title | LLM Braces: Straightening Out LLM Predictions with Relevant Sub-Updates |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2503.16334 |