Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains

Junhong Shen, Neil Tenenholtz, James Brian Hall, David Alvarez-Melis, Nicolo Fusi
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:44759-44773, 2024.

Abstract

Large Language Models (LLMs) have demonstrated remarkable proficiency in understanding and generating natural language. However, their capabilities wane in highly specialized domains underrepresented in the pretraining corpus, such as physical and biomedical sciences. This work explores how to repurpose general LLMs into effective task solvers for specialized domains. We introduce a novel, model-agnostic framework for learning custom input tags, which are parameterized as continuous vectors appended to the LLM’s embedding layer, to condition the LLM. We design two types of input tags: domain tags are used to delimit specialized representations (e.g., chemical formulas) and provide domain-relevant context; function tags are used to represent specific functions (e.g., predicting molecular properties) and compress function-solving instructions. We develop a three-stage protocol to learn these tags using auxiliary data and domain knowledge. By explicitly disentangling task domains from task functions, our method enables zero-shot generalization to unseen problems through diverse combinations of the input tags. It also boosts LLM’s performance in various specialized domains, such as predicting protein or chemical properties and modeling drug-target interactions, outperforming expert models tailored to these tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-shen24f, title = {Tag-{LLM}: Repurposing General-Purpose {LLM}s for Specialized Domains}, author = {Shen, Junhong and Tenenholtz, Neil and Hall, James Brian and Alvarez-Melis, David and Fusi, Nicolo}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {44759--44773}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/shen24f/shen24f.pdf}, url = {https://proceedings.mlr.press/v235/shen24f.html}, abstract = {Large Language Models (LLMs) have demonstrated remarkable proficiency in understanding and generating natural language. However, their capabilities wane in highly specialized domains underrepresented in the pretraining corpus, such as physical and biomedical sciences. This work explores how to repurpose general LLMs into effective task solvers for specialized domains. We introduce a novel, model-agnostic framework for learning custom input tags, which are parameterized as continuous vectors appended to the LLM’s embedding layer, to condition the LLM. We design two types of input tags: domain tags are used to delimit specialized representations (e.g., chemical formulas) and provide domain-relevant context; function tags are used to represent specific functions (e.g., predicting molecular properties) and compress function-solving instructions. We develop a three-stage protocol to learn these tags using auxiliary data and domain knowledge. By explicitly disentangling task domains from task functions, our method enables zero-shot generalization to unseen problems through diverse combinations of the input tags. It also boosts LLM’s performance in various specialized domains, such as predicting protein or chemical properties and modeling drug-target interactions, outperforming expert models tailored to these tasks.} }
Endnote
%0 Conference Paper %T Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains %A Junhong Shen %A Neil Tenenholtz %A James Brian Hall %A David Alvarez-Melis %A Nicolo Fusi %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-shen24f %I PMLR %P 44759--44773 %U https://proceedings.mlr.press/v235/shen24f.html %V 235 %X Large Language Models (LLMs) have demonstrated remarkable proficiency in understanding and generating natural language. However, their capabilities wane in highly specialized domains underrepresented in the pretraining corpus, such as physical and biomedical sciences. This work explores how to repurpose general LLMs into effective task solvers for specialized domains. We introduce a novel, model-agnostic framework for learning custom input tags, which are parameterized as continuous vectors appended to the LLM’s embedding layer, to condition the LLM. We design two types of input tags: domain tags are used to delimit specialized representations (e.g., chemical formulas) and provide domain-relevant context; function tags are used to represent specific functions (e.g., predicting molecular properties) and compress function-solving instructions. We develop a three-stage protocol to learn these tags using auxiliary data and domain knowledge. By explicitly disentangling task domains from task functions, our method enables zero-shot generalization to unseen problems through diverse combinations of the input tags. It also boosts LLM’s performance in various specialized domains, such as predicting protein or chemical properties and modeling drug-target interactions, outperforming expert models tailored to these tasks.
APA
Shen, J., Tenenholtz, N., Hall, J.B., Alvarez-Melis, D. & Fusi, N.. (2024). Tag-LLM: Repurposing General-Purpose LLMs for Specialized Domains. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:44759-44773 Available from https://proceedings.mlr.press/v235/shen24f.html.

Related Material