paint-brush
When Deductive Reasoning Fails: Contextual Ambiguities in AI Modelsby@cosmological
192 reads

When Deductive Reasoning Fails: Contextual Ambiguities in AI Models

by Cosmological thinking: time, space and universal causation
Cosmological thinking: time, space and universal causation  HackerNoon profile picture

Cosmological thinking: time, space and universal causation

@cosmological

From Big Bang's singularity to galaxies' cosmic dance the universe...

September 8th, 2024
Read on Terminal Reader
Read this story in a terminal
Print this story
Read this story w/o Javascript
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

Despite its effectiveness, the Natural Program-based deductive reasoning process has limitations, particularly in handling contextual ambiguities. One key failure case involves the term "pennies," where ChatGPT misinterprets it as a unit of currency instead of coins. Such ambiguities reveal challenges in the deductive verification process, limiting its ability to resolve contextual misunderstandings.
featured image - When Deductive Reasoning Fails: Contextual Ambiguities in AI Models
1x
Read by Dr. One voice-avatar

Listen to this story

Cosmological thinking: time, space and universal causation  HackerNoon profile picture
Cosmological thinking: time, space and universal causation

Cosmological thinking: time, space and universal causation

@cosmological

From Big Bang's singularity to galaxies' cosmic dance the universe unfolds its majestic tapestry of space and time.

Learn More
LEARN MORE ABOUT @COSMOLOGICAL'S
EXPERTISE AND PLACE ON THE INTERNET.
0-item

STORY’S CREDIBILITY

Academic Research Paper

Academic Research Paper

Part of HackerNoon's growing list of open-source research papers, promoting free access to academic material.

Authors:

(1) Zhan Ling, UC San Diego and equal contribution;

(2) Yunhao Fang, UC San Diego and equal contribution;

(3) Xuanlin Li, UC San Diego;

(4) Zhiao Huang, UC San Diego;

(5) Mingu Lee, Qualcomm AI Research and Qualcomm AI Research

(6) Roland Memisevic, Qualcomm AI Research;

(7) Hao Su, UC San Diego.

Abstract and Introduction

Related work

Motivation and Problem Formulation

Deductively Verifiable Chain-of-Thought Reasoning

Experiments

Limitations

Conclusion, Acknowledgements and References


A Deductive Verification with Vicuna Models

B More Discussion on Improvements of Deductive Verification Accuracy Versus Improvements on Final Answer Correctness

C More Details on Answer Extraction

D Prompts

E More Deductive Verification Examples

6 Limitations

While we have demonstrated the effectiveness of Natural Program-based deductive reasoning verification to enhance the trustworthiness and interpretability of reasoning steps and final answers, it is


Table 7: Ablation of different values of k ′ on the verification accuracy of reasoning chains using our Unanimity-Plurality Voting strategy. Experiments are performed on AddSub using GPT-3.5-turbo (ChatGPT).

Table 7: Ablation of different values of k ′ on the verification accuracy of reasoning chains using our Unanimity-Plurality Voting strategy. Experiments are performed on AddSub using GPT-3.5-turbo (ChatGPT).


Table 8: An example question with ambiguous wordings. The term "pennies" in this question can be interpreted as either a type of coin or a unit of currency. In this particular question, "pennies" is treated as a type of coin. However, the initial reasoning step by ChatGPT mistakenly treats "pennies" as a unit of currency, resulting in the conversion of all Melanie’s money into "pennies" (highlighted in red). Consequently, all subsequent reasoning steps follow this flawed logic, leading to an incorrect reasoning trace. Our deductive verification is not yet able to detect such errors.

Table 8: An example question with ambiguous wordings. The term "pennies" in this question can be interpreted as either a type of coin or a unit of currency. In this particular question, "pennies" is treated as a type of coin. However, the initial reasoning step by ChatGPT mistakenly treats "pennies" as a unit of currency, resulting in the conversion of all Melanie’s money into "pennies" (highlighted in red). Consequently, all subsequent reasoning steps follow this flawed logic, leading to an incorrect reasoning trace. Our deductive verification is not yet able to detect such errors.


important to acknowledge that our approach has limitations. In this section, we analyze a common source of failure cases to gain deeper insights into the behaviors of our approach. The failure case, as shown in Tab. 8, involves the ambiguous interpretation of the term “pennies,” which can be understood as either a type of coin or a unit of currency depending on the context. The ground truth answer interprets “pennies” as coins, while ChatGPT interprets it as a unit of currency. In this case, our deductive verification process is incapable of finding such misinterpretations. Contextual ambiguities like this are common in real-world scenarios, highlighting the current limitation of our approach.


This paper is available on arxiv under CC BY 4.0 DEED license.


L O A D I N G
. . . comments & more!

About Author

Cosmological thinking: time, space and universal causation  HackerNoon profile picture
Cosmological thinking: time, space and universal causation @cosmological
From Big Bang's singularity to galaxies' cosmic dance the universe unfolds its majestic tapestry of space and time.

TOPICS

THIS ARTICLE WAS FEATURED IN...

Permanent on Arweave
Read on Terminal Reader
Read this story in a terminal
 Terminal
Read this story w/o Javascript
Read this story w/o Javascript
 Lite
Also published here
X
X REMOVE AD