Yu-Xiong Wang, Ziyang Xie, Ziqi Pang, Yunze Man
12 of 16 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; 4 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 7 failures of our instrument; 4 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/2310.12973.json · badge: /api/ran/2310.12973.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 |
|---|---|---|
| ziqipang/lm4visualencoding | canonical (classifier label) classifier basis not recorded on the link |
12 ran · 4 no recorded run |
| Function | Status | Where it lives |
|---|---|---|
| add_weight_decay | Ran ran; our contract draft was wrong, not the code |
ziqipang/lm4visualencoding/image_classification/optimizer_utils.py code served (permissive licence) · get_code("e805e183bad0f027") How it ranRan 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.
|
| apply_rotary_emb | Ran ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT) |
ziqipang/lm4visualencoding/image_classification/models/llama.py pointer only (licence: MIT); the stored copy get_code serves, from meituan-automl/visionllama, is pointer only (licence: NONE) · get_code("b47d48e431b34acd") How it ranThe 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 meituan-automl/visionllama, and GitHub detected no licence in that repository 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. |
| build_transform | Ran ran; our contract draft was wrong, not the code |
ziqipang/lm4visualencoding/image_classification/datasets.py code served (permissive licence) · get_code("3bc74137ab36aa79") How it ranRan 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 Compose. 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.
|
| get_loss_scale_for_deepspeed | Ran ran, honoured the contract we drafted |
ziqipang/lm4visualencoding/video_understanding/engine_for_finetuning.py pointer only (licence: MIT); the stored copy get_code serves, from kangsankim07/videoicl, is pointer only (licence: NONE) · get_code("29416043c7035e7c") How it ranThe 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 kangsankim07/videoicl, and GitHub detected no licence in that repository 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. |
| get_mce_from_accuracy | Ran ran, honoured the contract we drafted |
ziqipang/lm4visualencoding/image_classification/utils.py code served (permissive licence) · get_code("a9e48df75d2e823e") How it ranRan 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 float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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.
|
| get_sinusoid_encoding_table | Ran ran, honoured the contract we drafted |
ziqipang/lm4visualencoding/video_understanding/modeling_finetune.py pointer only (licence: MIT); the stored copy get_code serves, from MCG-NJU/VideoMAE-Action-Detection, is pointer only (licence: NOASSERTION) · get_code("da651e3979a18f84") How it ranThe 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 MCG-NJU/VideoMAE-Action-Detection, 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. |
| optimizer_kwargs | Ran ran; our contract draft was wrong, not the code |
ziqipang/lm4visualencoding/image_classification/optimizer_utils.py pointer only (licence: MIT); the stored copy get_code serves, from Navantock/SE3Set, is pointer only (licence: NONE) · get_code("2cd1e282432041c8") How it ranThe 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 Navantock/SE3Set, and GitHub detected no licence in that repository 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. |
| precompute_freqs_cis | Ran ran, violated the contract we drafted |
ziqipang/lm4visualencoding/image_classification/models/llama.py pointer only (licence: MIT); the stored copy get_code serves, from meituan-automl/visionllama, is pointer only (licence: NONE) · get_code("14a84c2cbfebc413") How it ranThe 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 meituan-automl/visionllama, and GitHub detected no licence in that repository 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. |
| reshape_for_broadcast | Ran ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT) |
ziqipang/lm4visualencoding/image_classification/models/llama.py pointer only (licence: MIT); the stored copy get_code serves, from alpha-vllm/llama2-accessory, is pointer only (licence: NOASSERTION) · get_code("70bf6ebaafd266c4") How it ranThe 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 alpha-vllm/llama2-accessory, 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. |
| tensor_normalize | Ran ran, honoured the contract we drafted |
ziqipang/lm4visualencoding/video_understanding/kinetics.py pointer only (licence: MIT); the stored copy get_code serves, from kangsankim07/videoicl, is pointer only (licence: NONE) · get_code("14b3b13b2e7ee844") How it ranThe 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 kangsankim07/videoicl, and GitHub detected no licence in that repository 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. |
| train_class_batch | Ran ran; our contract draft was wrong, not the code |
ziqipang/lm4visualencoding/video_understanding/engine_for_finetuning.py pointer only (licence: MIT); the stored copy get_code serves, from kangsankim07/videoicl, is pointer only (licence: NONE) · get_code("c7976ea27edc377a") How it ranThe 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 kangsankim07/videoicl, and GitHub detected no licence in that repository 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. |
| unnormalize_image | Ran ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT) |
ziqipang/lm4visualencoding/image_classification/utils.py code served (permissive licence) · get_code("32208d3133397a32") How it ranRan 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 ndarray, ndim 3, dtype family float, shape (4, 5, 3) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape 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.
|
| PGDAttack | Not yet run unverified (harvested, no recorded run) |
ziqipang/lm4visualencoding/image_classification/engine.py code served (permissive licence) · get_code("58ab30e8c22165b6") How it ranno invocation recorded. |
| build_dataset | Not yet run unverified (harvested, no recorded run) |
ziqipang/lm4visualencoding/image_classification/datasets.py code served (permissive licence) · get_code("8b43480472330eaf") How it ranno invocation recorded. |
| clamp | Not yet run unverified (harvested, no recorded run) |
ziqipang/lm4visualencoding/image_classification/engine.py code served (permissive licence) · get_code("8a93e041134b597a") How it ranno invocation recorded. |
| patch_level_aug | Not yet run unverified (harvested, no recorded run) |
ziqipang/lm4visualencoding/image_classification/engine.py code served (permissive licence) · get_code("a57f32dae793402d") How it ranno invocation recorded. |
Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 16 are also present in other papers whose code we hold (10 with at least one paper dated earlier, 4 the same month, 16 later) and 0 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.
| Function | Also present in | Earliest-dated | Others named |
|---|---|---|---|
| reshape_for_broadcast | 58 other papers 6 dated earlier, 4 the same month, 47 later, 1 same year, order not determinable · 2 only as an assembly brick |
2023.emnlp-main.534 (2023 (year only)) Disentangling Transformer Language Models as Superposed Topic Models |
2303.09736 (2023-03) Dynamic Structure Pruning for Compressing CNNs 2307.09288 (2023-07) Llama 2: Open Foundation and Fine-Tuned Chat Models and 55 more not named here |
| apply_rotary_emb | 55 other papers 5 dated earlier, 4 the same month, 45 later, 1 same year, order not determinable · 1 only as an assembly brick |
2023.emnlp-main.534 (2023 (year only)) Disentangling Transformer Language Models as Superposed Topic Models |
2303.09736 (2023-03) Dynamic Structure Pruning for Compressing CNNs 2307.09288 (2023-07) Llama 2: Open Foundation and Fine-Tuned Chat Models and 52 more not named here |
| precompute_freqs_cis | 47 other papers 5 dated earlier, 4 the same month, 37 later, 1 same year, order not determinable · 1 only as an assembly brick |
2023.emnlp-main.534 (2023 (year only)) Disentangling Transformer Language Models as Superposed Topic Models |
2303.09736 (2023-03) Dynamic Structure Pruning for Compressing CNNs 2307.09288 (2023-07) Llama 2: Open Foundation and Fine-Tuned Chat Models and 44 more not named here |
| build_transform | 37 other papers 26 dated earlier, 8 later, 3 same year, order not determinable |
2104.05707 (2021-04) LocalViT: Analyzing Locality in Vision Transformers |
2104.13840 (2021-04) Twins: Revisiting the Design of Spatial Attention in Vision Transformers 2106.00515 (2021-06) KVT: k-NN Attention for Boosting Vision Transformers and 34 more not named here |
| clamp | 33 other papers 22 dated earlier, 1 the same month, 9 later, 1 same year, order not determinable · 1 only as an assembly brick |
1706.06083 (2017-06) Towards Deep Learning Models Resistant to Adversarial Attacks |
2001.03994 (2020-01) Fast is better than free: Revisiting adversarial training 2002.11569 (2020-02) Overfitting in adversarially robust deep learning and 30 more not named here |
| get_loss_scale_for_deepspeed | 15 other papers 5 dated earlier, 10 later |
2203.14415 (2022-03) Mugs: A Multi-Granular Self-Supervised Learning Framework |
2211.09120 (2022-11) AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoders 2212.10368 (2022-12) Masked Event Modeling: Self-Supervised Pretraining for Event Cameras and 12 more not named here |
| tensor_normalize | 15 other papers 5 dated earlier, 10 later |
2210.00788 (2022-10) Towards a Unified View on Visual Parameter-Efficient Transfer Learning |
2211.09120 (2022-11) AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoders 2307.02227 (2023-07) MAE-DFER: Efficient Masked Autoencoder for Self-supervised Dynamic Facial Expression Recognition and 12 more not named here |
| get_sinusoid_encoding_table | 13 other papers 6 dated earlier, 7 later · 4 only as an assembly brick |
2203.12119 (2022-03) Visual Prompt Tuning |
2203.12602 (2022-03) VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training 2211.10636 (2022-11) EVEREST: Efficient Masked Video Autoencoder by Removing Redundant Spatiotemporal Tokens and 10 more not named here |
| train_class_batch | 11 other papers 4 dated earlier, 7 later |
2203.14415 (2022-03) Mugs: A Multi-Granular Self-Supervised Learning Framework |
2211.09120 (2022-11) AdaMAE: Adaptive Masking for Efficient Spatiotemporal Learning with Masked Autoencoders 2212.10368 (2022-12) Masked Event Modeling: Self-Supervised Pretraining for Event Cameras and 8 more not named here |
| get_mce_from_accuracy | 4 other papers 2 dated earlier, 1 later, 1 same year, order not determinable |
2210.08823 (2022-10) Scaling & Shifting Your Features: A New Baseline for Efficient Model Tuning |
Guo_Improving_Robustness_of_Vision_Transformers_by_Reducing_Sensitivity_To_Patch_CVPR_2023_paper (2023 (year only)) Improving Robustness of Vision Transformers by Reducing Sensitivity to Patch Corruptions 2307.11342 (2023-07) Tuning Pre-trained Model via Moment Probing and 1 more not named here |
| PGDAttack | 2 other papers 1 later, 1 same year, order not determinable |
Guo_Improving_Robustness_of_Vision_Transformers_by_Reducing_Sensitivity_To_Patch_CVPR_2023_paper (2023 (year only)) Improving Robustness of Vision Transformers by Reducing Sensitivity to Patch Corruptions |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
| optimizer_kwargs | 2 other papers 2 later |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
2405.16511 (2024-05) SE3Set: Harnessing equivariant hypergraph neural networks for molecular representation learning |
| patch_level_aug | 2 other papers 1 later, 1 same year, order not determinable |
Guo_Improving_Robustness_of_Vision_Transformers_by_Reducing_Sensitivity_To_Patch_CVPR_2023_paper (2023 (year only)) Improving Robustness of Vision Transformers by Reducing Sensitivity to Patch Corruptions |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
| add_weight_decay | 1 other paper 1 later |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
|
| build_dataset | 1 other paper 1 later |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
|
| unnormalize_image | 1 other paper 1 later |
2403.17343 (2024-03) Residual-based Language Models are Free Boosters for Biomedical Imaging |
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.
This paper reveals that large language models (LLMs), despite being trained solely on text data, are surprisingly strong encoders for purely visual tasks in the absence of language. Even more intriguingly, this can be achieved by a simple yet previously overlooked strategy -employing a frozen transformer block from pretrained LLMs as a constituent encoder layer to directly process visual tokens. Our work pushes the boundaries of leveraging LLMs for computer vision tasks, significantly departing from conventional practices that typically necessitate a multimodal vision-language setup with associated language prompts, inputs, or outputs. We demonstrate that our approach consistently enhances performance across a diverse range of tasks, encompassing purely 2D and 3D visual recognition tasks (e.g., image and point cloud classification), temporal modeling tasks (e.g., action recognition), non-semantic tasks (e.g., motion forecasting), and multi-modal tasks (e.g., 2D/3D visual question answering and image-text retrieval). Such improvements are a general phenomenon, applicable to various types of LLMs (e.g., LLaMA and OPT) and different LLM transformer blocks. We additionally propose the information filtering hypothesis to explain the effectiveness of pre-trained LLMs in visual encoding -the pre-trained LLM transformer blocks discern informative visual tokens and further amplify their effect. This hypothesis is empirically supported by the observation that the feature activation, after training with LLM transformer blocks, exhibits a stronger focus on relevant regions. We hope that our work inspires new perspectives on utilizing LLMs and deepening our understanding of their underlying mechanisms.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2310.12973")
get_code_for_paper("2310.12973")
have("2310.12973")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2310.12973.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2310.12973)
Connect an agent — have() is free.