SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2510.16596 · ICLR · 2025

SHIELD: Suppressing Hallucinations In LVLM Encoders via Bias and Vulnerability Defense

Yun Fu, Yi Xu, Yitian Zhang, Yiyang Huang, Liang Shi

arXiv · PDF · Open in the Atlas

Code that ran

2 of 5 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; 2 failures of our instrument; 3 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/2510.16596.json · badge: /api/ran/2510.16596.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
hukcc/SHIELD community
Papers with Code archive
2 ran · 3 no recorded run
FunctionStatusWhere it lives
map_text_segments Ran
ran; our contract draft was wrong, not the code
hukcc/SHIELD/shield/wrapper.py
code served (permissive licence) · get_code("0e59ae7086d78972")
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, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) 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.

import torch

torch.manual_seed(0)

# Construct arguments for map_text_segments
# The function expects:
# - top_k_indices: tensor of indices into cap_tensor
# - cap_tensor: tensor with shape (batch, L_text, ...)
# - input_cap_tensor: tensor with shape (L_other, ...)

# Create top_k_indices - a 1D tensor of indices
top_k_indices = torch.tensor([0, 1, 2, 5, 6, 7], dtype=torch.long)

# Create cap_tensor with L_text=10 (second dimension)
# Shape: (batch_size, L_text, feature_dim)
cap_tensor = torch.randn(2, 10, 4)

# Create input_cap_tensor with L_other=8
# Shape: (L_other, feature_dim)
input_cap_tensor = torch.randn(8, 4)

ARGS = {
    "top_k_indices": top_k_indices,
    "cap_tensor": cap_tensor,
    "input_cap_tensor": input_cap_tensor,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
maximize_weight_difference Ran
ran; our contract draft was wrong, not the code
hukcc/SHIELD/shield/wrapper.py
code served (permissive licence) · get_code("ee7b487ea158a7a1")
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, through the batch API) and returned Tensor, ndim 1, dtype family float, shape (4,) 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.

import torch

torch.manual_seed(0)

# Create a small tensor with values that will demonstrate the weight stretching
weights = torch.tensor([0.2, 0.5, 0.3, 0.8], dtype=torch.float32)

# Define gamma parameter for power transformation
gamma = 1.0

ARGS = {
    "weights": weights,
    "gamma": gamma,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (4,) — shape (4,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
_shield_prepare_multimodal Not yet run
unverified (harvested, no recorded run)
hukcc/SHIELD/shield/wrapper.py
code served (permissive licence) · get_code("dd64e7c63010d212")
How it ran

no invocation recorded.

compute_shield_image_weights Not yet run
unverified (harvested, no recorded run)
hukcc/SHIELD/shield/wrapper.py
code served (permissive licence) · get_code("680779200dcc8389")
How it ran

no invocation recorded.

get_bias Not yet run
unverified (harvested, no recorded run)
hukcc/SHIELD/shield/wrapper.py
code served (permissive licence) · get_code("d20e0e91e0020b43")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 5 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

Large Vision-Language Models (LVLMs) excel in diverse cross-modal tasks. However, object hallucination, where models produce plausible but inaccurate object descriptions, remains a significant challenge. In contrast to previous work focusing on LLM components, this paper is the first to trace LVLM hallucinations to visual encoders and identifies three key issues: statistical bias, inherent bias, and vulnerability. To address these challenges, we propose SHIELD, a training-free framework that mitigates hallucinations through three strategies: re-weighting visual tokens to reduce statistical bias, introducing noise-derived tokens to counter inherent bias, and applying adversarial attacks with contrastive decoding to address vulnerability. Experiments demonstrate that SHIELD effectively mitigates object hallucinations across diverse benchmarks and LVLM families. Moreover, SHIELD achieves strong performance on the general LVLM benchmark, highlighting its broad applicability. Code is available at https: //github.com/hukcc/SHIELD.

For agents

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

get_harvested_code_for_paper("2510.16596")
get_code_for_paper("2510.16596")
have("2510.16596")

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

curl https://syntology.ai/api/ran/2510.16596.json

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

[![Syntology run record](https://syntology.ai/api/ran/2510.16596.svg)](https://syntology.ai/paper/2510.16596)

Connect an agent — have() is free.