Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction | Paper Citation Record

Observation e27eee67-71f9-4100-8bcb-157b5ef51e24 · outbound

This paper cites LLaVA-OneVision: Easy Visual Task Transfer.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction LLaVA-OneVision: Easy Visual Task Transfer

Reference 1

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.494370Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 73daf702-45dc-4ba5-9b8c-c00c0f426859 · outbound

This paper cites Qwen3-VL Technical Report.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Qwen3-VL Technical Report

Reference 2

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.499723Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 581d4fef-c86b-4679-892c-e199b0988514 · outbound

This paper cites InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models

Reference 3

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.503957Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 690036a4-7e49-4628-8a91-54aefacf29ac · outbound

This paper cites Visual Instruction Tuning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Visual Instruction Tuning

Reference 4

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.508431Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 073cf827-4cbb-4837-ab8f-b899bd3e0298 · outbound

This paper cites Learning transferable visual models from natural language supervision.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Learning transferable visual models from natural language supervision

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.513081Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 66e1153a-e3fe-4ab7-a84d-c0a1ec0e8aba · outbound

This paper cites Visual cot: Advancing multi-modal language models with a comprehensive dataset and benchmark for chain-of-thought reasoning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Visual cot: Advancing multi-modal language models with a comprehensive dataset and benchmark for chain-of-thought reasoning

Reference 6

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.517459Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 48223edc-b902-4f01-9d60-bb73da0595d2 · outbound

This paper cites An empirical anal- ysis on spatial reasoning capabilities of large multimodal models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction An empirical anal- ysis on spatial reasoning capabilities of large multimodal models

Reference 7

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.522320Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9dcd2796-7c6f-470c-b83f-a84dce81d3b9 · outbound

This paper cites Why are visually-grounded language models bad at image classification? Advances in Neural Information Processing Systems, 37:51727–51753, 2024.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Why are visually-grounded language models bad at image classification? Advances in Neural Information Processing Systems, 37:51727–51753, 2024

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.537516Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 56769107-c26a-4d69-8ea9-0a62b49efece · outbound

This paper cites Question aware vision transformer for multimodal reasoning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Question aware vision transformer for multimodal reasoning

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.521970Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation d29ae9be-8ff3-4050-9b9a-60a890622529 · outbound

This paper cites Chameleon: Mixed-Modal Early-Fusion Foundation Models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Chameleon: Mixed-Modal Early-Fusion Foundation Models

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.534077Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8111b51c-97ea-491e-bf8a-5ac7e5ac0c30 · outbound

This paper cites Show-o: One single transformer to unify multimodal understanding and generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Show-o: One single transformer to unify multimodal understanding and generation

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.505930Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 1b064072-c6b8-46a6-a37c-2e788fcd5d72 · outbound

This paper cites Transfusion: Predict the next token and diffuse images with one multi-modal model.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Transfusion: Predict the next token and diffuse images with one multi-modal model

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.490249Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 5357a682-5d79-41d5-9b67-aa6be7d681e1 · outbound

This paper cites Emerging Properties in Unified Multimodal Pretraining.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Emerging Properties in Unified Multimodal Pretraining

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.553581Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3f0b369f-d2e4-4afc-9563-01b08ee21b37 · outbound

This paper cites Mammothmoda2: A unified ar-diffusion framework for multimodal understanding and generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mammothmoda2: A unified ar-diffusion framework for multimodal understanding and generation

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.558435Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 569f7263-26fc-45b3-8025-46d8129b94ac · outbound

This paper cites Lance: Unified Multimodal Modeling by Multi-Task Synergy.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Lance: Unified Multimodal Modeling by Multi-Task Synergy

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.562911Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 63003f27-2aa1-4654-9c49-363a64cad2c1 · outbound

This paper cites Multimodal learning with next-token prediction for large multimodal models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Multimodal learning with next-token prediction for large multimodal models

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.567839Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 84770594-92dd-4d11-8867-6f6eccdd453c · outbound

This paper cites Janus: Decoupling visual encoding for unified multimodal understanding and generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Janus: Decoupling visual encoding for unified multimodal understanding and generation

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.572476Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5da4b6b3-7fb0-4f5a-befe-3a99babd2c83 · outbound

This paper cites Mmada: Multimodal large diffusion language models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mmada: Multimodal large diffusion language models

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.447949Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 75ccea02-af2c-4f5b-a6b5-5904a0f2c0b2 · outbound

This paper cites Longcat-next: Lexicalizing modalities as discrete tokens.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Longcat-next: Lexicalizing modalities as discrete tokens

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.581554Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 64c0448f-797f-402e-b382-d4c47a23a458 · outbound

This paper cites Next-embedding prediction makes strong vision learners.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Next-embedding prediction makes strong vision learners

Reference 20

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.585968Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 37c21e1c-d592-4b6d-ab4d-9da81a50ac3f · outbound

This paper cites Unihetero: Could generation enhance understanding for vision-language-model at large data scale? arXiv preprint arXiv:2512.23512, 2025.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Unihetero: Could generation enhance understanding for vision-language-model at large data scale? arXiv preprint arXiv:2512.23512, 2025

Reference 21

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.590538Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 942b4361-6ee5-44ce-ba8e-a87358e57908 · outbound

This paper cites UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation

Reference 22

Resolution: verified exact; local_arxiv, observed 2026-08-16T00:17:34.447253Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 0c30f67c-517b-4927-81ba-45a7ce43b7b8 · outbound

This paper cites Uni-x: Mitigating modality conflict with a two-end-separated architecture for unified multimodal models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Uni-x: Mitigating modality conflict with a two-end-separated architecture for unified multimodal models

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.599748Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 6985abff-97f7-4ef9-9bb6-0f31241ac41d · outbound

This paper cites Mixture-of-transformers: A sparse and scalable architecture for multi-modal foundation models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mixture-of-transformers: A sparse and scalable architecture for multi-modal foundation models

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.431635Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 1dfc456c-42b0-4591-ae33-cf1c8703e850 · outbound

This paper cites Qwen3 Technical Report.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Qwen3 Technical Report

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.608967Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ef9f9ac-56f1-4129-9351-ddfa7291dc70 · outbound

This paper cites Adam: A Method for Stochastic Optimization.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Adam: A Method for Stochastic Optimization

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.613869Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d32e69c6-17cd-44c4-abf9-8a2e7f63c81b · outbound

This paper cites Zero: Memory optimizations toward training trillion parameter models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Zero: Memory optimizations toward training trillion parameter models

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.415500Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 41ae2848-0ec8-47b1-90a8-5f6a04382004 · outbound

This paper cites Mme: A comprehensive evaluation benchmark for multimodal large language models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mme: A comprehensive evaluation benchmark for multimodal large language models

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.399745Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 183cc7c9-3aab-49f4-9be4-b059d9ce5cfd · outbound

This paper cites Mmmu: A massive multi-discipline multimodal understanding and reasoning benchmark for expert agi.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mmmu: A massive multi-discipline multimodal understanding and reasoning benchmark for expert agi

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.384039Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation f66b54ef-2f8e-4fa1-b366-dae59ce38d49 · outbound

This paper cites Blink: Multimodal large language models can see but not perceive.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Blink: Multimodal large language models can see but not perceive

Reference 30

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.632306Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d1930387-009c-43db-90c4-b86a88be5003 · outbound

This paper cites Charxiv: Charting gaps in realistic chart understanding in multimodal llms.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Charxiv: Charting gaps in realistic chart understanding in multimodal llms

Reference 31

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.636732Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 39e60b28-eb04-4718-9a89-3abff8768d58 · outbound

This paper cites Dynamath: A dynamic visual benchmark for evaluating mathematical reasoning robustness of vision language models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Dynamath: A dynamic visual benchmark for evaluating mathematical reasoning robustness of vision language models

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.641183Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a74c4a37-3aca-4d5d-94d0-28c8a50ffbdb · outbound

This paper cites Measuring multimodal mathematical reasoning with math-vision dataset.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Measuring multimodal mathematical reasoning with math-vision dataset

Reference 33

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.645477Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 320d399e-1b4b-4016-ac2b-82b8f1efb5be · outbound

This paper cites Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mathvista: Evaluating mathematical reasoning of foundation models in visual contexts

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.326990Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation f06d5710-f9c5-4b5f-a98d-e20707970c57 · outbound

This paper cites LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts

Reference 35

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.654244Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 02f09eb4-7996-4dfb-af6e-2b535747e291 · outbound

This paper cites Visulogic: A benchmark for evalu- ating visual reasoning in multi-modal large language models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Visulogic: A benchmark for evalu- ating visual reasoning in multi-modal large language models

Reference 36

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.311942Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 26c84fe8-d53e-442a-a4e8-85a94ad09106 · outbound

This paper cites Teaching CLIP to Count to Ten.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Teaching CLIP to Count to Ten

Reference 37

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.662073Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 1804ee86-6b67-4b94-947f-7787318f7452 · outbound

This paper cites PaliGemma: A versatile 3B VLM for transfer.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction PaliGemma: A versatile 3B VLM for transfer

Reference 38

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.665902Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation def81ece-2147-47a0-bae5-2907a8f1f1a4 · outbound

This paper cites Cambrian-1: A fully open, vision-centric exploration of multimodal llms.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Cambrian-1: A fully open, vision-centric exploration of multimodal llms

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.296526Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 40f3269a-697b-44c9-a810-275fa14c1ae3 · outbound

This paper cites Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal llms in video analysis.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Video-mme: The first-ever comprehensive evaluation benchmark of multi-modal llms in video analysis

Reference 40

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.280336Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 314f5c6e-7cb5-418f-b2df-3c51bcb1741d · outbound

This paper cites Mvbench: A comprehensive multi-modal video understanding benchmark.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Mvbench: A comprehensive multi-modal video understanding benchmark

Reference 41

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.677995Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 2cbd7d36-a227-4539-97e0-6a8033974b9a · outbound

This paper cites VisuLogic: A Benchmark for Evaluating Visual Reasoning in Multi-modal Large Language Models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction VisuLogic: A Benchmark for Evaluating Visual Reasoning in Multi-modal Large Language Models

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.682022Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation cd0019bf-cc06-4fde-a865-57ac6f7ea711 · outbound

This paper cites Thyme: Think beyond images.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Thyme: Think beyond images

Reference 43

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.254442Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 523eabbf-f825-472c-b160-5d5f89a617c0 · outbound

This paper cites Improved baselines with visual instruction tuning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Improved baselines with visual instruction tuning

Reference 44

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.691642Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 83854f49-131d-44c5-bbf6-29f72bce05df · outbound

This paper cites Qwen2.5-VL Technical Report, February 2025.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Qwen2.5-VL Technical Report, February 2025

Reference 45

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.229128Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation e64aa5f8-a4cc-45e2-8d69-c4de94a0fee5 · outbound

This paper cites Metamorph: Multimodal understanding and generation via instruction tuning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Metamorph: Multimodal understanding and generation via instruction tuning

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.212857Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 6a154796-6cdf-4c11-a87a-e2d00e90aabb · outbound

This paper cites Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.704691Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c858fa4b-9750-475f-82ff-832482d94d18 · outbound

This paper cites Show-o2: Improved native unified multimodal models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Show-o2: Improved native unified multimodal models

Reference 48

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.196673Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation afd85e1c-5dd4-4d8f-8ef3-6c2d91e10f02 · outbound

This paper cites Cheers: Decoupling patch details from semantic representations enables unified multimodal comprehension and generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Cheers: Decoupling patch details from semantic representations enables unified multimodal comprehension and generation

Reference 49

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.713410Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 50dfeed4-2f57-42eb-a738-4a3adbb969c6 · outbound

This paper cites Coyo-700m: Image-text pair dataset.https://github.com/kakaobrain/coyo-dataset, 2022.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Coyo-700m: Image-text pair dataset.https://github.com/kakaobrain/coyo-dataset, 2022

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.718062Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 48670c1e-7796-4337-84f5-53a46388c1c1 · outbound

This paper cites Imagenet: A large-scale hierarchical image database.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Imagenet: A large-scale hierarchical image database

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.722979Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5ecf47df-c7d9-4053-b9f6-444cd7a69bcc · outbound

This paper cites Gen- eration and comprehension of unambiguous object descriptions.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Gen- eration and comprehension of unambiguous object descriptions

Reference 52

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.161417Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 3091c69c-ee9c-4e96-a108-d3f6d8b20ad8 · outbound

This paper cites Referitgame: Referring to objects 23 in photographs of natural scenes.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Referitgame: Referring to objects 23 in photographs of natural scenes

Reference 53

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.145538Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 26584f7a-e5c5-430d-8e1e-9bb1d93aa0af · outbound

This paper cites ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.737344Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c70cbff7-5db0-4c6d-b223-3480478740f4 · outbound

This paper cites an unresolved cited work.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Unresolved cited work

Reference 55

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.742241Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 905e03c1-049e-4f64-a7bc-241d4ddf8505 · outbound

This paper cites Flamingo: a visual language model for few-shot learning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Flamingo: a visual language model for few-shot learning

Reference 56

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.747386Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9268cac0-a2b7-4868-b137-d12d81434bcb · outbound

This paper cites Instructblip: Towards general-purpose vision-language models with instruction tuning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Instructblip: Towards general-purpose vision-language models with instruction tuning

Reference 57

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.111944Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 84d506e9-63f8-4616-8f32-61a9a384bdfe · outbound

This paper cites LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.758009Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 806d6f5a-3a59-438d-b1ec-c1e99e70adca · outbound

This paper cites SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.762555Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bd887cbb-d5da-4d59-9661-9de455213db5 · outbound

This paper cites gpt-5-system-card, 2025.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction gpt-5-system-card, 2025

Reference 60

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.097932Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation aacebb07-2ed8-446c-91cb-8ecf55737892 · outbound

This paper cites URL https://blog.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction URL https://blog

Reference 61

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.082321Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 53b7ce97-9450-476f-be32-e5a9b92933bf · outbound

This paper cites Gemini 3 flash: frontier intelligence built for speed, 2025.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Gemini 3 flash: frontier intelligence built for speed, 2025

Reference 62

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.067638Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation fd929ecf-1804-4e14-a388-44982ba0013b · outbound

This paper cites VGR: Visual grounded reasoning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction VGR: Visual grounded reasoning

Reference 63

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.052990Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation e58da99f-15e6-4f75-ae40-28dbe6609a75 · outbound

This paper cites Explain Before You Answer: A Survey on Compositional Visual Reasoning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Explain Before You Answer: A Survey on Compositional Visual Reasoning

Reference 64

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.785540Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8f707088-def0-4a1a-9541-a31fde45abf9 · outbound

This paper cites Video-XL-Pro: Reconstructive Token Compression for Extremely Long Video Understanding.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Video-XL-Pro: Reconstructive Token Compression for Extremely Long Video Understanding

Reference 65

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.789852Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ac6385c7-46e4-4ff2-ae91-b502cbe09e08 · outbound

This paper cites Video-xl: Extra-long vision language model for hour-scale video understanding.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Video-xl: Extra-long vision language model for hour-scale video understanding

Reference 66

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.036455Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation a929b0c3-f6b3-4cd3-b19b-0091995455d4 · outbound

This paper cites Videochat-flash: Hierarchical compression for long-context video modeling.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Videochat-flash: Hierarchical compression for long-context video modeling

Reference 67

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.020711Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation a4a286f7-2de5-4dce-8344-2dd178faa383 · outbound

This paper cites Reconstructive visual instruction tuning.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Reconstructive visual instruction tuning

Reference 68

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:35.004875Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation e779dc83-cfa0-4390-b4fb-f2934a771cda · outbound

This paper cites Autoregressive semantic visual reconstruction helps vlms understand better.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Autoregressive semantic visual reconstruction helps vlms understand better

Reference 69

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.989707Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 0dae9be8-18b4-462f-aedf-d247e85fa217 · outbound

This paper cites Generation enhances understanding in unified multimodal models via multi-representation generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Generation enhances understanding in unified multimodal models via multi-representation generation

Reference 70

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.975035Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation b8ab97a0-4659-420f-bbd2-40479f921270 · outbound

This paper cites BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-16T00:17:33.916989Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f820c2a0-3031-4f0e-87b8-b377bd0734ae · outbound

This paper cites LMFusion: Adapting pretrained language models for multimodal generation.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction LMFusion: Adapting pretrained language models for multimodal generation

Reference 72

Resolution: verified exact; raw_fallback, observed 2026-08-16T00:17:34.087176Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 95c572ab-905d-4dd6-bc6f-629fc15d2719 · outbound

This paper cites - Row 1, Fig.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction - Row 1, Fig

Reference 74

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.960789Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 3b16e616-1bc7-4869-aa16-07fc00ab01c8 · outbound

This paper cites All figures share the same two-shape overlapping composition with consistent contour and inner-line connectivity.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction All figures share the same two-shape overlapping composition with consistent contour and inner-line connectivity

Reference 75

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.947379Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 2c55ad56-d9a3-48ba-9d36-a7298d51ff94 · outbound

This paper cites Option A is a triangle pair with a mismatched shape; Option B is a triangle pair with inconsistent line count; Option C is a quadrilateral but with a different stacking style.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Option A is a triangle pair with a mismatched shape; Option B is a triangle pair with inconsistent line count; Option C is a quadrilateral but with a different stacking style

Reference 76

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.931691Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 64359b35-023e-4c33-8ad6-d23ce8ab9a83 · outbound

This paper cites A bipartite graph is a graph whose vertices can be divided into two disjoint sets such that every edge connects a vertex in one set to a vertex in the other set.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction A bipartite graph is a graph whose vertices can be divided into two disjoint sets such that every edge connects a vertex in one set to a vertex in the other set

Reference 77

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.915799Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation ace91903-90da-48b4-8a18-a0dfcaaadbf3 · outbound

This paper cites If there are no odd-length cycles, the graph is bipartite.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction If there are no odd-length cycles, the graph is bipartite

Reference 78

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.899750Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation fed92e39-9f71-4eb9-b408-a516c7389120 · outbound

This paper cites This is because we can color one set of vertices with one color and the other set with a different color, ensuring that no two adjacent vertices share the same color.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction This is because we can color one set of vertices with one color and the other set with a different color, ensuring that no two adjacent vertices share the same color

Reference 79

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.883082Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 7a933e72-25a3-440a-84da-7f5ef6be6931 · outbound

This paper cites </think> The final answer is2.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction </think> The final answer is2

Reference 80

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.866845Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 1cb9d5ad-aa2a-48ce-adbc-690eaec86790 · outbound

This paper cites There are two visible wooden poles supporting the tree.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction There are two visible wooden poles supporting the tree

Reference 81

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.850621Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation 349c59f0-13ea-4c7c-ab7d-5e1d5209b7af · outbound

This paper cites There are no additional wooden poles visible in the background or elsewhere in the image.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction There are no additional wooden poles visible in the background or elsewhere in the image

Reference 82

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.833597Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation efe688c9-dee5-47cf-8bef-abf21d9a764e · outbound

This paper cites Given this analysis, the correct answer is: **C.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Given this analysis, the correct answer is: **C

Reference 83

Resolution: verified fuzzy; raw_fallback, observed 2026-08-16T00:17:34.817360Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Observation ff1816c0-e15b-4c2c-99e6-b95510212187 · outbound

This paper cites an unresolved cited work.

Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction Unresolved cited work

Reference 2025

Resolution: unresolved; raw_fallback, observed 2026-08-16T00:17:35.474150Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-23T06:30:58.430688+00:00.

Coverage vector

Reference resolution

Outbound references

Pith citing papers