카테고리 없음

my paper

jinuklee 2024. 7. 22. 21:24

hallucination, toxic content, faulty code

 

A problem-solving process is a sequence of iterative stages within a human group (Bransford & Stein, 1993).

 

this research aims to solve multi-step reasoning taskusing multi-agent framework 

Despite the rapid advancement of LLM, its shortcomings of LLMs in areas such as mathematical computations, reasoning, and factuality interrupt application lacks in especially multi-step reasoning task

but existing multi-agent framework 

 

This is due to the overly long token sequences and the interleaved text-image data format, which limit performance.

and their cooperative behaviours with other AI are still not explored well.

 

prior research has focused on (IOA, evoagent, 

 

multi-agent framework where strong agents supervise weak models demonstrate

 

Large language models (LLMs) are expected to respond accurately but often exhibit deficient reasoning or generate

hallucinatory content

 

On the other hand, solving complex reasoning problems requires more in-depth, deliberative and logical thinking steps, i.e., the “System 2" mode [15]. Taking solving math word problems as an example, any incorrect intermediate reasoning step (e.g., calculation errors, mis-interpretations) can potentially lead to incorrect final answers

 

In multi-agent frameworks (Zeng et al., 2022; Zhuge et al., 2023), several LLMs take on different roles (Li et al., 2023; Park et al., 2023; Qian et al., 2023; Wu et al., 2023), to communicate in natural language and collectively solve a given task. This approach often outperforms single agents, exploiting the specialization (Hong et al., 2023) of various LLM agents. Unfortunately, it also leads to increasingly different and disparate code bases that require a lot of human engineering to define prompting schemes and the workflow of agents.

 

From the perspective of multi-agent system robustness, any unexpected error or response can propagate to the whole system, causing a series of cascading effects if not handled properly

 

In the process of completing complex tasks, interactive reasoning among agents within a static MACNET requires strategical traversal to establish an orderly interaction sequence

 

Soft 프롬프트

Hard prompt is highly sensitive to word choice, where even minor changes can significantly impact generation quality. To address these limitations, soft prompt methods use continuous, trainable vector embeddings, offering more flexible and fine-grained control without altering the underlying model parameters.

 

task decomposition

self feedback

computation graph optimization

discussion

 

fall short

suboptimal

challenges

streamline

outline

underline

underscore

facilitate

alleviate

devised

denote, indicate (e.g denoted as, represented by

demonstrate

remarkable

non-trivial

capability

prior

tailored

leverage

prescribe

employ

address 

exhibit

desired result

assess

empower

is designed to

elicit

decompose

scalable

reveal that

as shown in

Given

serve as

reliable

uncover

bypass

extensive

emperical

propose

as illustrated in

primarily

substantial

seamlessly integrate

the potential

application in

baseline

refers to (mean)

enable 

ensure

compatible with

as follows: (1)

signify

assign

induce 

comprehensive

workflow

conduct

tackle

calibrated

discrepancy

investigate

charactarized by

implementation

modular

robust

concurrent

varying (e.g varying levels of expertise)

travelplanning 데이터셋

Quality 데이터셋

 

power of law와 structure에 관해 MACNET