CELI: Controller-Embedded Language Model Interactions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wagner, Jan-Samuel, DeCaprio, Dave, Raja, Abishek Chiffon Muthu, Holman, Jonathan M., Brady, Lauren K., Cheung, Sky C., Barzekar, Hosein, Yang, Eric, Martinez II, Mark Anthony, Soong, David, Sridhar, Sriram, Si, Han, Higgs, Brandon W., Hamadeh, Hisham, Ogden, Scott
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916445176201216
author Wagner, Jan-Samuel
DeCaprio, Dave
Raja, Abishek Chiffon Muthu
Holman, Jonathan M.
Brady, Lauren K.
Cheung, Sky C.
Barzekar, Hosein
Yang, Eric
Martinez II, Mark Anthony
Soong, David
Sridhar, Sriram
Si, Han
Higgs, Brandon W.
Hamadeh, Hisham
Ogden, Scott
author_facet Wagner, Jan-Samuel
DeCaprio, Dave
Raja, Abishek Chiffon Muthu
Holman, Jonathan M.
Brady, Lauren K.
Cheung, Sky C.
Barzekar, Hosein
Yang, Eric
Martinez II, Mark Anthony
Soong, David
Sridhar, Sriram
Si, Han
Higgs, Brandon W.
Hamadeh, Hisham
Ogden, Scott
contents We introduce Controller-Embedded Language Model Interactions (CELI), a framework that integrates control logic directly within language model (LM) prompts, facilitating complex, multi-stage task execution. CELI addresses limitations of existing prompt engineering and workflow optimization techniques by embedding control logic directly within the operational context of language models, enabling dynamic adaptation to evolving task requirements. Our framework transfers control from the traditional programming execution environment to the LMs, allowing them to autonomously manage computational workflows while maintaining seamless interaction with external systems and functions. CELI supports arbitrary function calls with variable arguments, bridging the gap between LMs' adaptive reasoning capabilities and conventional software paradigms' structured control mechanisms. To evaluate CELI's versatility and effectiveness, we conducted case studies in two distinct domains: code generation (HumanEval benchmark) and multi-stage content generation (Wikipedia-style articles). The results demonstrate notable performance improvements across a range of domains. CELI achieved a 4.9 percentage point improvement over the best reported score of the baseline GPT-4 model on the HumanEval code generation benchmark. In multi-stage content generation, 94.4% of CELI-produced Wikipedia-style articles met or exceeded first draft quality when optimally configured, with 44.4% achieving high quality. These outcomes underscore CELI's potential for optimizing AI-driven workflows across diverse computational domains.
format Preprint
id arxiv_https___arxiv_org_abs_2410_14627
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle CELI: Controller-Embedded Language Model Interactions
Wagner, Jan-Samuel
DeCaprio, Dave
Raja, Abishek Chiffon Muthu
Holman, Jonathan M.
Brady, Lauren K.
Cheung, Sky C.
Barzekar, Hosein
Yang, Eric
Martinez II, Mark Anthony
Soong, David
Sridhar, Sriram
Si, Han
Higgs, Brandon W.
Hamadeh, Hisham
Ogden, Scott
Software Engineering
Artificial Intelligence
Computation and Language
68T50, 68Q32, 68N19
I.2.6; I.2.7; D.2.2
We introduce Controller-Embedded Language Model Interactions (CELI), a framework that integrates control logic directly within language model (LM) prompts, facilitating complex, multi-stage task execution. CELI addresses limitations of existing prompt engineering and workflow optimization techniques by embedding control logic directly within the operational context of language models, enabling dynamic adaptation to evolving task requirements. Our framework transfers control from the traditional programming execution environment to the LMs, allowing them to autonomously manage computational workflows while maintaining seamless interaction with external systems and functions. CELI supports arbitrary function calls with variable arguments, bridging the gap between LMs' adaptive reasoning capabilities and conventional software paradigms' structured control mechanisms. To evaluate CELI's versatility and effectiveness, we conducted case studies in two distinct domains: code generation (HumanEval benchmark) and multi-stage content generation (Wikipedia-style articles). The results demonstrate notable performance improvements across a range of domains. CELI achieved a 4.9 percentage point improvement over the best reported score of the baseline GPT-4 model on the HumanEval code generation benchmark. In multi-stage content generation, 94.4% of CELI-produced Wikipedia-style articles met or exceeded first draft quality when optimally configured, with 44.4% achieving high quality. These outcomes underscore CELI's potential for optimizing AI-driven workflows across diverse computational domains.
title CELI: Controller-Embedded Language Model Interactions
topic Software Engineering
Artificial Intelligence
Computation and Language
68T50, 68Q32, 68N19
I.2.6; I.2.7; D.2.2
url https://arxiv.org/abs/2410.14627