Papers › Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens

Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens

20 Jun 2025arXiv:2506.17218archive 2025-07-28

Zeyuan Yang, Xueyang Yu, Delin Chen, Maohao Shen, Chuang Gan

Vision-language models (VLMs) excel at multimodal understanding, yet their text-only decoding forces them to verbalize visual reasoning, limiting performance on tasks that demand visual imagination. Recent attempts train VLMs to render explicit images, but the heavy image-generation pre-training often hinders the reasoning ability. Inspired by the way humans reason with mental imagery-the internal construction and manipulation of visual cues-we investigate whether VLMs can reason through interleaved multimodal trajectories without producing explicit images. To this end, we present a Machine Mental Imagery framework, dubbed as Mirage, which augments VLM decoding with latent visual tokens alongside ordinary text. Concretely, whenever the model chooses to ``think visually'', it recasts its hidden states as next tokens, thereby continuing a multimodal trajectory without generating pixel-level images. Begin by supervising the latent tokens through distillation from ground-truth image embeddings, we then switch to text-only supervision to make the latent trajectory align tightly with the task objective. A subsequent reinforcement learning stage further enhances the multimodal reasoning capability. Experiments on diverse benchmarks demonstrate that Mirage unlocks stronger multimodal reasoning without explicit image generation.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2506.17218")

Code

Syntology Ran 0 of 17 code samples harvested from 1 repository linked to this paper; 17 have no recorded run.

By repository: official repository: 17 samples from 1 repository, 0 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

umass-embodied-agi/mirage officialmentioned in papermentioned on GitHubjax report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

17 samples harvested; 0 ran; 0 honoured the contract we drafted; 17 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

17unverified

Licence: 17 of the 17 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from umass-embodied-agi/mirage. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

build_tip_message umass-embodied-agi/mirage/transformers/utils/deprecate_models.py official repository unverified no licence file found · pointer only · ab306165fa74656c · report
extract_model_info umass-embodied-agi/mirage/transformers/utils/deprecate_models.py official repository unverified no licence file found · pointer only · 54b7273383fcfc2d · report
get_architectures_from_config_class umass-embodied-agi/mirage/transformers/utils/create_dummy_models.py official repository unverified no licence file found · pointer only · b176efb26f72e507 · report
get_cased_name umass-embodied-agi/mirage/transformers/utils/modular_model_converter.py official repository unverified no licence file found · pointer only · 8d6745e2692cf03d · report
get_config_class_from_processor_class umass-embodied-agi/mirage/transformers/utils/create_dummy_models.py official repository unverified no licence file found · pointer only · 3d999b95009c7537 · report
get_list_of_repo_model_paths umass-embodied-agi/mirage/transformers/utils/models_to_deprecate.py official repository unverified no licence file found · pointer only · 938fc7c2422ef91d · report
get_model_doc_path umass-embodied-agi/mirage/transformers/utils/deprecate_models.py official repository unverified no licence file found · pointer only · 051d790d9b197d9d · report
get_module_source_from_name umass-embodied-agi/mirage/transformers/utils/modular_model_converter.py official repository unverified no licence file found · pointer only · 029be6c1f552852b · report
get_new_python_files_between_commits umass-embodied-agi/mirage/transformers/utils/pr_slow_ci_models.py official repository unverified no licence file found · pointer only · 0f78c7037b35a137 · report
get_processor_types_from_config_class umass-embodied-agi/mirage/transformers/utils/create_dummy_models.py official repository unverified no licence file found · pointer only · 0c0b81b081919d8e · report
load_jsonl_dataset umass-embodied-agi/mirage/src/utils.py official repository unverified no licence file found · pointer only · 0fbb3cf4415efb41 · report
multiple_input_images_preprocess_function umass-embodied-agi/mirage/src/task.py official repository unverified no licence file found · pointer only · 01f09967a9ef73ab · report
place_input_image umass-embodied-agi/mirage/src/utils.py official repository unverified no licence file found · pointer only · 67012056b57c9945 · report
place_output_image umass-embodied-agi/mirage/src/utils.py official repository unverified no licence file found · pointer only · a6823237c70a51ca · report
preserve_case_replace umass-embodied-agi/mirage/transformers/utils/modular_model_converter.py official repository unverified no licence file found · pointer only · d646d0a64800a8fe · report
single_input_image_preprocess_function umass-embodied-agi/mirage/src/task.py official repository unverified no licence file found · pointer only · b76973ab8137eb01 · report
single_input_image_test_preprocess_function umass-embodied-agi/mirage/src/task.py official repository unverified no licence file found · pointer only · c9aa7df020dc5885 · report

Tasks

Image GenerationMultimodal ReasoningVisual Reasoning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

ALIGN

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections