# Forgery Reasoning on TFR CIS Cross-Image-domain Source: [Forgery Reasoning on TFR CIS Cross-Image-domain]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:01:10.506Z Evaluation of model performance on the Forgery Reasoning task using the TFR CIS Cross-Image-domain dataset, measured by Reasoning Score (Avg. Cosine/Rouge-L/BLEU). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Forgery Reasoning | | Dataset (including any stated split) | TFR CIS Cross-Image-domain | | Metric | Reasoning Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Reasoning Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Reasoning Score | Result date | Report paper | | --- | --- | --- | --- | | TextShield-R1 | 56.5 | Not recorded | 699d30aa6017c4460ffab609 | | Qwen2.5-VL-7B | 35.7 | Not recorded | 699d30aa6017c4460ffab609 | | SIDA\* | 35.7 | Not recorded | 699d30aa6017c4460ffab609 | | FakeShield\* | 35.6 | Not recorded | 699d30aa6017c4460ffab609 | | InternVL3-8B | 33.8 | Not recorded | 699d30aa6017c4460ffab609 | | Qwen2.5-VL-3B | 33.6 | Not recorded | 699d30aa6017c4460ffab609 | | MiniCPM\_V\_2.6 | 32.3 | Not recorded | 699d30aa6017c4460ffab609 | | InternVL3-2B | 32 | Not recorded | 699d30aa6017c4460ffab609 | | SIDA | 29.4 | Not recorded | 699d30aa6017c4460ffab609 | | FakeShield | 29.2 | Not recorded | 699d30aa6017c4460ffab609 | | GPT4o | 24.3 | Not recorded | 699d30aa6017c4460ffab609 | | InternVL3-8B | 20.3 | Not recorded | 699d30aa6017c4460ffab609 | | Qwen2.5-VL-7B | 17.6 | Not recorded | 699d30aa6017c4460ffab609 | | Qwen2.5-VL-3B | 15.3 | Not recorded | 699d30aa6017c4460ffab609 | | InternVL3-2B | 9.9 | Not recorded | 699d30aa6017c4460ffab609 | | MiniCPM\_V\_2.6 | 3.2 | Not recorded | 699d30aa6017c4460ffab609 | ### TextShield-R1 Reasoning Score: 56.5. Result ID: 699e56415698665ec4166304. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Reasoning Score: 35.7. Result ID: 699e56415698665ec41662ff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### SIDA\* Reasoning Score: 35.7. Result ID: 699e56415698665ec4166303. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images","Base MLLM":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### FakeShield\* Reasoning Score: 35.6. Result ID: 699e56415698665ec4166301. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images","Base MLLM":"Qwen2.5-VL-7B"}. This record does not assert independent reproduction. ### InternVL3-8B Reasoning Score: 33.8. Result ID: 699e56415698665ec41662fd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B Reasoning Score: 33.6. Result ID: 699e56415698665ec41662fe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### MiniCPM\_V\_2.6 Reasoning Score: 32.3. Result ID: 699e56415698665ec41662fb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### InternVL3-2B Reasoning Score: 32. Result ID: 699e56415698665ec41662fc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### SIDA Reasoning Score: 29.4. Result ID: 699e56415698665ec4166302. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### FakeShield Reasoning Score: 29.2. Result ID: 699e56415698665ec4166300. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"Full training set images"}. This record does not assert independent reproduction. ### GPT4o Reasoning Score: 24.3. Result ID: 699e56415698665ec41662f5. [Reporting paper]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. ### InternVL3-8B Reasoning Score: 20.3. Result ID: 699e56415698665ec41662f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Reasoning Score: 17.6. Result ID: 699e56415698665ec41662fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. ### Qwen2.5-VL-3B Reasoning Score: 15.3. Result ID: 699e56415698665ec41662f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. ### InternVL3-2B Reasoning Score: 9.9. Result ID: 699e56415698665ec41662f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. ### MiniCPM\_V\_2.6 Reasoning Score: 3.2. Result ID: 699e56415698665ec41662f6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Fine-tuning":"None"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.