SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2606.28044 · 2026

A Tree-of-Thoughts Hybrid Approach for Legal Case Judgement Summarization using LLMs

Iiser Kolkata, Aniket Deroy, India Ghosh

arXiv · PDF · Open in the Atlas

Code that ran

8 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 8 failures of our instrument; 0 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2606.28044.json · badge: /api/ran/2606.28044.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
Yale-LILY/SummEval canonical (classifier label)
classifier basis: census
8 ran · 0 no recorded run
FunctionStatusWhere it lives
cook_refs Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/cider_utils.py
code served (permissive licence) · get_code("cdec5e1ad686ef23")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

from collections import defaultdict

# Set random seed for reproducibility
import torch
torch.manual_seed(0)

# Define ARGS dict with all parameters for cook_refs
ARGS = {
    "refs": [
        "the quick brown fox jumps over the lazy dog",
        "a fast brown fox leaps over a lazy dog",
        "the swift fox jumps over the lazy canine"
    ],
    "n": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
six, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
cook_test Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/cider_utils.py
code served (permissive licence) · get_code("542ce8d78e319d1e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned defaultdict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from collections import defaultdict

torch.manual_seed(0)

ARGS = {
    "test": "the quick brown fox jumps over the lazy dog",
    "n": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
six, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type defaultdict — a defaultdict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a defaultdict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
dec Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/meteor_metric.py
pointer only (licence: MIT); the stored copy get_code serves, from Maluuba/nlg-eval, is pointer only (licence: NOASSERTION) · get_code("ba9348aac71a3f72")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Maluuba/nlg-eval, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

enc Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/meteor_metric.py
pointer only (licence: MIT); the stored copy get_code serves, from Maluuba/nlg-eval, is pointer only (licence: NOASSERTION) · get_code("eaab50b5d9fea58a")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Maluuba/nlg-eval, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

find_ngrams Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/data_stats_metric.py
code served (permissive licence) · get_code("15011c2229e5cd8b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned zip. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# find_ngrams takes an input_list and n
# It creates n sliding windows over the list and zips them together
# This is a simple function that doesn't require complex objects

# Example: input_list could be a list of tokens, words, or numbers
# n is the size of n-grams to extract

input_list = [1, 2, 3, 4, 5, 6, 7, 8]
n = 3

ARGS = {
    "input_list": input_list,
    "n": n
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type zip — a zip output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a zip output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalize Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/data_stats_utils.py
code served (permissive licence) · get_code("5a954296aa8dfe56")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Analyze the function:
# - tokens: iterable that will be looped over
# - case: boolean flag (default False)
# - Each element t in tokens is converted to string and optionally lowercased
# - Returns a list of processed strings

# Create test tokens - can be strings, numbers, or any objects with __str__
tokens = ["HELLO", "World", 123, "TEST"]

ARGS = {
    "tokens": tokens,
    "case": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
parse_story_file Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/data_processing/pair_data.py
code served (permissive licence) · get_code("314e4a6841248541")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

def parse_story_file(content):
    """
    Remove article highlights and unnecessary white characters.
    """
    content_raw = content.split('@highlight')[0]
    content = ' '.join(filter(None, [x.strip() for x in content_raw.split('\n')]))
    return content

# Construct valid arguments for parse_story_file
# The function expects a string 'content' parameter
sample_content = """This is the first paragraph of the story.
This is the second paragraph.

This is the third paragraph with some spacing.
@highlight
This is a highlight that should be removed.
@highlight
Another highlight."""

ARGS = {
    'content': sample_content
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
precook Ran
ran; our contract draft was wrong, not the code
Yale-LILY/SummEval/evaluation/summ_eval/cider_utils.py
code served (permissive licence) · get_code("eaa4133f5d797afc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned defaultdict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

from collections import defaultdict

# precook takes a string 's' and an integer 'n' for ngram size
# The function splits the string into words and creates ngrams of sizes 1 to n
# It returns a defaultdict of ngram tuple counts

ARGS = {
    "s": "the quick brown fox jumps over the lazy dog",
    "n": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
six — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type defaultdict — a defaultdict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a defaultdict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 0 the same month, 0 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
cook_refs 25 other papers
25 dated earlier
1411.5726 (2014-11)
CIDEr: Consensus-based Image Description Evaluation
1502.03044 (2015-02)
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
1503.01817 (2015-03)
YFCC100M: The New Data in Multimedia Research
and 22 more not named here
dec 4 other papers
4 dated earlier
1706.09799 (2017-06)
Relevance of Unsupervised Metrics in Task-Oriented Dialogue for Evaluating Natural Language Generation
2109.08365 (2021-09)
CodeQA: A Question Answering Dataset for Source Code Comprehension
2022.findings-naacl.160 (2022 (year only))
Learning Structural Information for Syntax-Controlled Paraphrase Generation
and 1 more not named here
enc 4 other papers
4 dated earlier
1706.09799 (2017-06)
Relevance of Unsupervised Metrics in Task-Oriented Dialogue for Evaluating Natural Language Generation
2109.08365 (2021-09)
CodeQA: A Question Answering Dataset for Source Code Comprehension
2022.findings-naacl.160 (2022 (year only))
Learning Structural Information for Syntax-Controlled Paraphrase Generation
and 1 more not named here
precook 4 other papers
4 dated earlier
1611.09571 (2016-11)
Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model
1804.08771 (2018-04)
A Call for Clarity in Reporting BLEU Scores
1912.08226 (2019-12)
Meshed-Memory Transformer for Image Captioning
and 1 more not named here
cook_test 2 other papers
2 dated earlier
1611.09571 (2016-11)
Predicting Human Eye Fixations via an LSTM-based Saliency Attentive Model
1912.08226 (2019-12)
Meshed-Memory Transformer for Image Captioning
parse_story_file 1 other paper
1 dated earlier
2020.emnlp-main.750 (2020 (year only))
Evaluating the Factual Consistency of Abstractive Text Summarization

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

In recent times, Large Language Models (LLMs) are increasingly being used for legal case judgement summarization. Most prior works have tried traditional extractive and abstractive summarization of case judgements. However, hybrid or extractive-abstractive techniques have not been explored much. In this work, we propose a novel tree-of-thoughts inspired extractive-abstractive summarization approach for legal judgement summarization. We conduct experiments using two popular LLMs, DeepSeek and LLama, and compare among extractive, abstractive and extractive-abstractive summarization. Our experiments show that the proposed extractiveabstractive prompt provides better summaries compared to other types of LLM prompts.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2606.28044")
get_code_for_paper("2606.28044")
have("2606.28044")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2606.28044.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2606.28044.svg)](https://syntology.ai/paper/2606.28044)

Connect an agent — have() is free.