Jun Xu, Jun Zhao, Jintao Fei, Wen Tang, Zhuoyun Liu, Zhenhong Yang, Shichao Quan
9 of 10 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; 3 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 5 failures of our instrument; 1 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/2606.17062.json · badge: /api/ran/2606.17062.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.
| Repository | Role | Ran |
|---|---|---|
| jdh-algo/RadSEM | community Papers with Code archive |
9 ran · 1 no recorded run |
| Function | Status | Where it lives |
|---|---|---|
| _Dinic | Ran ran on a synthesized input · constructed an object; no result was computed |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("94a0f279d6a53355") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _allocated_match_count | Ran ran, honoured the contract we drafted |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("8a5448269b496388") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _class_score | Ran ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT) |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("c320337f1d082ac5") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _clean_sentence | Ran ran; our contract draft was wrong, not the code |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("b007cffe0fac66d5") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _collect_weighted_pairs | Ran ran; our contract draft was wrong, not the code |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("c14a672a6c0c2a7f") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _deduplicate_pair_edges | Ran ran; our contract draft was wrong, not the code |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("3edd17b8dafac67b") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _pair_weight | Ran ran, honoured the contract we drafted |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("017e9ec74afc4b41") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _unique_pair_weights | Ran ran, honoured the contract we drafted |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("41f6fad5e30dcf89") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| _unmatched_counts | Ran ran; our contract draft was wrong, not the code |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("1fef4c265bddcc6a") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. |
| calculate_score | Not yet run unverified (harvested, no recorded run) |
jdh-algo/RadSEM/step/step3.py pointer only (licence: NONE) · get_code("29b27aedb8af8403") How it ranThe body of this sample is withheld: GitHub detected no licence in this paper's repository, jdh-algo/RadSEM when it was checked. 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. No invocation recorded. |
Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 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 10 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.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Radiology reports are built from compact clinical statements, where a single negation, laterality marker, or normal-abnormal polarity can change the clinical meaning. Existing metrics usually approach report evaluation from one of three directions: they compare textual resemblance, make holistic error judgments, or align extracted fine-grained findings. These views are useful, but they do not always systematically answer a local clinical question: does each clinical fact in a generated report agree with the reference facts, especially when abnormalities are missed or altered, or when statements that cannot be true together are nevertheless treated as similar? We propose RadSEM (Radiology Sentence-Level Evaluation Metric), a constrained LLM-assisted metric for reference-based scoring of radiology Findings. RadSEM first rewrites reference and generated reports into ordered atomic finding sentences, each expressing one site-finding proposition after non-finding content has been removed. It then performs contradiction-constrained many-to-many semantic matching between these units, so that lexically similar but clinically incompatible statements, such as "no effusion" and "effusion," do not receive credit, while clinically compatible granularity differences can still be represented as pair-level partial matches. Each accepted pair is weighted according to part-whole relationships and abnormal-detail coverage, and a deterministic scoring stage aggregates the pair weights under one unit of matched-credit capacity per atomic sentence, together with unmatched sentences, into an abnormal-focused weighted F1 score. In this design, the LLM supports structured rewriting and local semantic alignment rather than acting as an opaque global judge. RadSEM focuses on Findings rather than Impressions to reduce confounding between image-reading agreement and context-dependent diagnostic synthesis. We assess RadSEM with SSREE, a controlled monotonicity stress test built from 2,448 de-identified, English-translated real reports and five corruption levels (12,240 scoring instances), and with a 599-triplet synonym/antonym rewrite subset. In these controlled experiments, RadSEM obtains Kendall τ b of 0.957, all-pairs concordance of 97.8%, adjacent-level concordance of 95.0%, and strict five-level ordering for 81.9% of reports on SSREE; on the rewrite subset, it prefers synonymous over antonymous rewrites in 597 of 599 cases (99.67%). These results suggest that explicit statement units and contradiction-constrained matching can make automatic report scoring more interpretable and less vulnerable to clinically misleading surface similarity. The code is available at https://github.com/jdh-algo/RadSEM.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2606.17062")
get_code_for_paper("2606.17062")
have("2606.17062")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2606.17062.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2606.17062)
Connect an agent — have() is free.