SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.09898 · ACL · 2023

Balancing Lexical and Semantic Quality in Abstractive Summarization

Yong Choi, Jeewoo Sul

arXiv · PDF · Open in the Atlas

Code that ran

1 of 1 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2305.09898.json · badge: /api/ran/2305.09898.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
jeewoo1025/BalSum community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
BalSum Ran
ran on a synthesized input · constructed an object; no result was computed
jeewoo1025/BalSum/model.py
code served (permissive licence) · get_code("3f771a6f0e886148")
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 an instance of its own class, BalSum: the run constructed the object and did not call it, so no result was computed. 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 torch import nn
import torch.nn.functional as F
from transformers import RobertaModel
from unittest.mock import Mock, MagicMock

torch.manual_seed(0)

# Create a mock RobertaModel that doesn't require downloading
class MockRobertaModel(nn.Module):
    def __init__(self, hidden_size=4):
        super().__init__()
        self.hidden_size = hidden_size
        self.embeddings = nn.Embedding(1000, hidden_size)
        self.encoder_layer = nn.TransformerEncoderLayer(
            d_model=hidden_size, nhead=1, dim_feedforward=8, batch_first=True
        )
        self.encoder = nn.TransformerEncoder(self.encoder_layer, num_layers=1)

    def forward(self, input_ids, attention_mask=None):
        embedded = self.embeddings(input_ids)
        output = self.encoder(embedded, src_key_padding_mask=~attention_mask.bool())
        return (output,)

    @classmethod
    def from_pretrained(cls, model_name, cache_dir=None):
        return cls(hidden_size=4)

# Monkey patch to avoid actual download
original_from_pretrained = RobertaModel.from_pretrained
RobertaModel.from_pretrained = MockRobertaModel.from_pretrained

# Create the encoder instance
encoder = "roberta-base"
pad_token_id = 0
cls_token_id = 101
hidden_size = 4
temp = 0.1
gpuid = 0

# Create test data
batch_size = 2
seq_len = 10
sum_num = 3
token_dim = 4

# text_id: (batch_size, seq_len) - document tokens
text_id = torch.randint(1, 100, (batch_size, seq_len))
text_id[0, 2] = cls_token_id  # Add cls tokens
text_id[0, 5] = cls_token_id
text_id[1, 3] = cls_token_id
text_id[1, 7] = cls_token_id

# candidate_id: (batch_size, sum_num, seq_len) - candidate summary tokens
candidate_id = torch.randint(1, 100, (batch_size, sum_num, seq_len))

# neg_id: (batch_size, sum_num, seq_len) - negative summary tokens
neg_id = torch.randint(1, 100, (batch_size, sum_num, seq_len))

ARGS = {
    'encoder': encoder,
    'pad_token_id': pad_token_id,
    'cls_token_id': cls_token_id,
    'hidden_size': hidden_size,
    'temp': temp,
    'gpuid': gpuid,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type BalSum — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a BalSum 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 1 samples lifted from this paper's linked repositories, 1 carry a function identity and 0 do not. Of the 1 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 1 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.

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

An important problem of the sequence-tosequence neural models widely used in abstractive summarization is exposure bias. To alleviate this problem, re-ranking systems have been applied in recent years. Despite some performance improvements, this approach remains underexplored. Previous works have mostly specified the rank through the ROUGE score and aligned candidate summaries, but there can be quite a large gap between the lexical overlap metric and semantic similarity. In this paper, we propose a novel training method in which a re-ranker balances the lexical and semantic quality. We further newly define false positives in ranking and present a strategy to reduce their influence. Experiments on the CNN/DailyMail and XSum datasets show that our method can estimate the meaning of summaries without seriously degrading the lexical aspect. More specifically, it achieves an 89.67 BERTScore on the CNN/DailyMail dataset, reaching new state-of-the-art performance. Our code is publicly available at https: //github.com/jeewoo1025/BalSum.

For agents

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

get_harvested_code_for_paper("2305.09898")
get_code_for_paper("2305.09898")
have("2305.09898")

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

curl https://syntology.ai/api/ran/2305.09898.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.09898.svg)](https://syntology.ai/paper/2305.09898)

Connect an agent — have() is free.