SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.19862 · 2026

IsoCLIP: Decomposing CLIP Projectors for Efficient Intra-modal Alignment

Bartłomiej Twardowski, Joost Van De Weijer, Andrew Bagdanov, Dipam Goswami, Simone Magistri, Marco Mistretta

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 (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure 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/2603.19862.json · badge: /api/ran/2603.19862.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
simomagi/IsoCLIP community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
apply_iso Ran
ran; our contract draft was wrong, not the code
simomagi/IsoCLIP/src/retrieval.py
code served (permissive licence) · get_code("7333d4ab17ff863e")
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) and returned tuple. 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

torch.manual_seed(0)

# W_text and W_image are weight matrices used in SVD decomposition
# W_image.T @ W_text produces Psi matrix
# The function expects these to be 2D tensors (matrices)
# Using distinct dimensions: W_text is (4, 3), W_image is (4, 5)
# This ensures U will be (4, 3), V will be (3, 3) after SVD

W_text = torch.randn(4, 3)
W_image = torch.randn(4, 5)

# iso_ktop and iso_kbottom are integers controlling which singular components to filter
# They must satisfy: iso_ktop + iso_kbottom < r (where r is min dimension of Psi)
# Psi will be (5, 3), so r = 3
# We use iso_ktop=0, iso_kbottom=1 to keep the middle component

iso_ktop = 0
iso_kbottom = 1

ARGS = {
    'W_text': W_text,
    'W_image': W_image,
    'iso_ktop': iso_ktop,
    'iso_kbottom': iso_kbottom,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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

Vision-Language Models like CLIP are extensively used for inter-modal tasks which involve both visual and text modalities. However, when the individual modality encoders are applied to inherently intra-modal tasks like image-to-image retrieval, their performance suffers from the intra-modal misalignment. In this paper we study intra-modal misalignment in CLIP with a focus on the role of the projectors that map pre-projection image and text embeddings into the shared embedding space. By analyzing the form of the cosine similarity applied to projected features, and its interaction with the contrastive CLIP loss, we show that there is an inter-modal operator responsible for aligning the two modalities during training, and a second, intra-modal operator that only enforces intra-modal normalization but does nothing to promote intra-modal alignment. Via spectral analysis of the intermodal operator, we identify an approximately isotropic subspace in which the two modalities are well-aligned, as well as anisotropic directions specific to each modality. We demonstrate that this aligned subspace can be directly obtained from the projector weights and that removing the anisotropic directions improves intra-modal alignment. Our experiments on intra-modal retrieval and classification benchmarks show that our training-free method reduces intra-modal misalignment, greatly lowers latency, and outperforms existing approaches across multiple pretrained CLIP-like models. The code is publicly available at: https://github.com/simomagi/IsoCLIP.

For agents

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

get_harvested_code_for_paper("2603.19862")
get_code_for_paper("2603.19862")
have("2603.19862")

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

curl https://syntology.ai/api/ran/2603.19862.json

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

[![Syntology run record](https://syntology.ai/api/ran/2603.19862.svg)](https://syntology.ai/paper/2603.19862)

Connect an agent — have() is free.