cat-llm 0.0.73__tar.gz → 0.0.74__tar.gz

{cat_llm-0.0.73 → cat_llm-0.0.74}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cat-llm
-Version: 0.0.73
+Version: 0.0.74
 Summary: A tool for categorizing text data and images using LLMs and vision models
 Project-URL: Documentation, https://github.com/chrissoria/cat-llm#readme
 Project-URL: Issues, https://github.com/chrissoria/cat-llm/issues
@@ -200,6 +200,7 @@ Processes each text response individually, assigning one or more categories from
 - `safety` (bool, default=False): Enable safety checks on responses and saves to CSV at each API call step
 - `to_csv` (bool, default=False): Whether to save results to CSV
 - `chain_of_verification` (bool, default=False): Enable Chain-of-Verification prompting technique for improved accuracy
+- `chain_of_thought` (bool, default=False): Enable Chain-of-Thought prompting technique for improved accuracy
 - `step_back_prompt` (bool, default=False): Enable step-back prompting to analyze higher-level context before classification
 - `context_prompt` (bool, default=False): Add expert role and behavioral guidelines to the prompt
 - `filename` (str, default="categorized_data.csv"): Filename for CSV output

{cat_llm-0.0.73 → cat_llm-0.0.74}/README.md

@@ -171,6 +171,7 @@ Processes each text response individually, assigning one or more categories from
 - `safety` (bool, default=False): Enable safety checks on responses and saves to CSV at each API call step
 - `to_csv` (bool, default=False): Whether to save results to CSV
 - `chain_of_verification` (bool, default=False): Enable Chain-of-Verification prompting technique for improved accuracy
+- `chain_of_thought` (bool, default=False): Enable Chain-of-Thought prompting technique for improved accuracy
 - `step_back_prompt` (bool, default=False): Enable step-back prompting to analyze higher-level context before classification
 - `context_prompt` (bool, default=False): Add expert role and behavioral guidelines to the prompt
 - `filename` (str, default="categorized_data.csv"): Filename for CSV output

{cat_llm-0.0.73 → cat_llm-0.0.74}/src/catllm/__about__.py

@@ -1,7 +1,7 @@
 # SPDX-FileCopyrightText: 2025-present Christopher Soria <chrissoria@berkeley.edu>
 #
 # SPDX-License-Identifier: MIT
-__version__ = "0.0.73"
+__version__ = "0.0.74"
 __author__ = "Chris Soria"
 __email__ = "chrissoria@berkeley.edu"
 __title__ = "cat-llm"

{cat_llm-0.0.73 → cat_llm-0.0.74}/src/catllm/text_functions.py

@@ -260,6 +260,7 @@ def multi_class(
     safety = False,
     to_csv = False,
     chain_of_verification = False,
+    chain_of_thought = True,
     step_back_prompt = False,
     context_prompt = False,
     filename = "categorized_data.csv",
@@ -397,12 +398,27 @@ def multi_class(
             extracted_jsons.append(default_json)
             #print(f"Skipped NaN input.")
         else:
+            if chain_of_thought:
+                prompt = f"""{survey_question_context}
 
-            prompt = f"""{survey_question_context} \
-            Categorize this survey response "{response}" into the following categories that apply: \
-            {categories_str}
-            {examples_text}
-            Provide your work in JSON format where the number belonging to each category is the key and a 1 if the category is present and a 0 if it is not present as key values."""
+                Categorize this survey response "{response}" into the following categories that apply: 
+                {categories_str}
+
+                Let's think step by step:
+                1. First, identify the main themes mentioned in the response
+                2. Then, match each theme to the relevant categories
+                3. Finally, assign 1 to matching categories and 0 to non-matching categories
+
+                {examples_text}
+
+                Provide your reasoning for each category, then provide your final answer in JSON format where the number belonging to each category is the key and a 1 if the category is present and a 0 if it is not present as key values."""
+            else:
+
+                prompt = f"""{survey_question_context} \
+                Categorize this survey response "{response}" into the following categories that apply: \
+                {categories_str}
+                {examples_text}
+                Provide your work in JSON format where the number belonging to each category is the key and a 1 if the category is present and a 0 if it is not present as key values."""
 
             if context_prompt:
                 context = """You are an expert researcher in survey data categorization.