SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2605.12960 · 2026

DiM 3 : Bridging Multilingual and Multimodal Models via Direction-and Magnitude-Aware Merging

Shi Feng, Mengjie Zhao, Hinrich Schütze, Ercong Nie, Mingyang Wang, Daling Wang, Xiaocui Yang, Yongkang Liu, Zijing Wang

arXiv · PDF · Open in the Atlas

Code that ran

9 of 9 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; 3 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 5 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/2605.12960.json · badge: /api/ran/2605.12960.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
wzj1718/DiM3 canonical (classifier label)
classifier basis: census
9 ran · 0 no recorded run
FunctionStatusWhere it lives
act Ran
ran; our contract draft was wrong, not the code
wzj1718/DiM3/merge/merging_methods/pcb_merging.py
pointer only (licence: NONE) · get_code("7e60730f07a2532a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

add_ptm_to_tv Ran
ran; our contract draft was wrong, not the code
wzj1718/DiM3/merge/merging_methods/ties_merging_utils.py
pointer only (licence: NONE) · get_code("860bf587f2fbf4e0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

check_state_dicts_equal Ran
ran, violated the contract we drafted
wzj1718/DiM3/merge/merging_methods/ties_merging_utils.py
pointer only (licence: NONE) · get_code("5a7ea2d1b13f410f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

clamp Ran
ran; our contract draft was wrong, not the code
wzj1718/DiM3/merge/merging_methods/pcb_merging.py
pointer only (licence: NONE) · get_code("76418510bf7df06d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

flatten_ckpt_into_vec Ran
ran, honoured the contract we drafted
wzj1718/DiM3/merge/merging_methods/utils.py
pointer only (licence: NONE) · get_code("c8856c78385f817a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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_task_vector Ran
ran, honoured the contract we drafted
wzj1718/DiM3/merge/merging_methods/utils.py
pointer only (licence: NONE) · get_code("fd399befaee47739")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

normalize Ran
ran; our contract draft was wrong, not the code
wzj1718/DiM3/merge/merging_methods/pcb_merging.py
pointer only (licence: NONE) · get_code("657972dcf33a3093")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

select_trainable_params Ran
ran; our contract draft was wrong, not the code
wzj1718/DiM3/merge/merging_methods/utils.py
pointer only (licence: NONE) · get_code("39ddec4bec43fc72")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

state_dict_to_vector Ran
ran, honoured the contract we drafted
wzj1718/DiM3/merge/merging_methods/ties_merging_utils.py
pointer only (licence: NONE); the stored copy get_code serves, from thunlp-mt/adamms, is pointer only (licence: NONE) · get_code("72391ca34e9a8cfe")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, wzj1718/DiM3 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.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 0 later) and 7 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
state_dict_to_vector 3 other papers
3 dated earlier
2405.17461 (2024-05)
EMR-MERGING: Tuning-Free High-Performance Model Merging
2410.14389 (2024-10)
SurgeryV2: Bridging the Gap Between Model Merging and Multi-Task Learning with Deep Representation Surgery
2503.23733 (2025-03)
AdaMMS: Model Merging for Heterogeneous Multimodal Large Language Models with Unsupervised Coefficient Optimization
add_ptm_to_tv 2 other papers
2 dated earlier
2311.13171 (2023-11)
ComPEFT: Compression for Communicating Parameter Efficient Updates via Sparsification and Quantization
2405.17461 (2024-05)
EMR-MERGING: Tuning-Free High-Performance Model Merging

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

Towards more general and human-like intelligence, large language models should seamlessly integrate both multilingual and multimodal capabilities; however, extending an existing multimodal model to many languages typically requires expensive multilingual multimodal data construction and repeated end-to-end retraining. We study a training-free alternative: injecting multilingual capability into an existing multimodal model by composing residual updates in the shared language model backbone. The key challenge is that multilingual and multimodal updates are heterogeneous, reflecting different functional roles in the shared model. To address this, we propose Directionand Magnitude-aware Multilingual Multimodal merging (DiM 3 ), which selectively composes the two updates at each parameter dimension while preserving the original vision encoder and multimodal projector. Experiments on multilingual benchmarks in both text-only and vision-language settings, covering 57 languages across LLaVA-and Qwen-based backbones, show that DiM 3 consistently outperforms existing merging baselines, substantially improves multilingual performance over the original multimodal model, and remains competitive with dedicated multilingual multimodal fine-tuning while largely retaining general multimodal ability. We further show that DiM 3 can be directly applied to already trained multilingual multimodal models and still yield additional gains. Further interpretability analysis shows that DiM 3 primarily reshapes intermediate-layer semantic representations, strengthening cross-lingual alignment under both textonly and multimodal inputs while preserving higher-layer task-sensitive structure. Our repository is on https://github.com/wzj1718/DiM3.

For agents

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

get_harvested_code_for_paper("2605.12960")
get_code_for_paper("2605.12960")
have("2605.12960")

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

curl https://syntology.ai/api/ran/2605.12960.json

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

[![Syntology run record](https://syntology.ai/api/ran/2605.12960.svg)](https://syntology.ai/paper/2605.12960)

Connect an agent — have() is free.