Paper Citation Record · LEDGER

Improving MLLM Training Efficiency via Stage-Aware Sparsity

As of 11 August 2026, this Paper Citation Record lists 34 of 34 outbound references and 1 inbound Pith citation observation for arXiv:2509.18150.

A citation records a reference. It does not transfer a finding from one paper to another.

pith.paper-citation-record.v1
2509.18150 v2

Coverage vector

measured 34 of 34 reference resolution

Typed states for the displayed outbound observations.

Source: paper_references, paper_reference_links, observed 2026-05-21T22:00:45.377456Z

measured 35 of 35 standing notices

One-hop event checks from named stored sources.

Source: scholarly_work_events, retraction_status_cache, observed 2026-08-11T06:34:44.6726+00:00

measured 1 of 1 inbound itemization

Pith citing papers itemized under the disclosed page cap.

Source: paper_references, paper_reference_links, observed 2026-05-21T22:00:45.377456Z

measured 0 of 1 external citation measurements

A source-named dated measurement, never combined with another source.

Source: pith, observed 2026-05-21T22:04:24.215885Z

Reference resolution

34 of 34 outbound references displayed

verified exact5
verified fuzzy28
unresolved1
parse uncertain0
malformed identifier0
metadata mismatch0

External citation measurements

No source-named external measurement is stored.

Outbound references

Observation 4e1fb56b-e447-4dbf-a374-e88b0fa0a53f · outbound

This paper cites an unresolved cited work.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Unresolved cited work

Reference 1

Resolution: unresolved; raw_fallback, observed 2026-05-21T22:04:24.922738Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation f9bba53d-bd7d-4c60-861f-b3b74aaf6722 · outbound

This paper cites Improving MLLM Training Efficiency via Stage-Aware Sparsity.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Improving MLLM Training Efficiency via Stage-Aware Sparsity

Reference 2

Resolution: verified exact; local_arxiv, observed 2026-05-21T22:04:24.217936Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 3f96adc6-b2fc-4153-9716-725467d35a82 · outbound

This paper cites We also illustrate the ablation studies in Sec.

Improving MLLM Training Efficiency via Stage-Aware Sparsity We also illustrate the ablation studies in Sec

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.929212Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 73a4119d-11c0-467a-8acd-13aecc428af3 · outbound

This paper cites STS reduces the computational overhead by using two components: VTC and LDS, which can compress redundant visual inputs and dynamically skip unnecessary decoder layers respectively.

Improving MLLM Training Efficiency via Stage-Aware Sparsity STS reduces the computational overhead by using two components: VTC and LDS, which can compress redundant visual inputs and dynamically skip unnecessary decoder layers respectively

Reference 4

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.925913Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 5a40032d-362b-4be9-971a-033545dc5970 · outbound

This paper cites Glm: General language model pretraining with autoregressive blank infilling.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Glm: General language model pretraining with autoregressive blank infilling

Reference 5

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.932043Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 7742a0af-d60f-44b6-9d72-d164dfc0a478 · outbound

This paper cites Language models are few-shot learners.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Language models are few-shot learners

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.935309Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 68119adc-d53e-40d4-ae33-dd2a82c2b49c · outbound

This paper cites Llama: Open and efficient foundation language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Llama: Open and efficient foundation language models

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.889872Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 8ca377b3-dd7a-40e9-a2e9-1ef4a7badb25 · outbound

This paper cites Qwen2.5 technical report.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Qwen2.5 technical report

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.914494Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 953397f2-46d9-4dee-bbbb-17009d76bc62 · outbound

This paper cites Blip- 2: Bootstrapping language-image pre-training with frozen im- age encoders and large language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Blip- 2: Bootstrapping language-image pre-training with frozen im- age encoders and large language models

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.918130Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation a63decf4-66c0-472a-b131-029eb635cb4f · outbound

This paper cites In- ternvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks.

Improving MLLM Training Efficiency via Stage-Aware Sparsity In- ternvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks

Reference 10

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.928898Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 67dded0d-5605-4e11-baa1-86c8acfd2846 · outbound

This paper cites Qwen-vl: A versatile vision-language model for understand- ing, localization, text reading, and beyond.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Qwen-vl: A versatile vision-language model for understand- ing, localization, text reading, and beyond

Reference 11

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.925275Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 50eda628-0ad0-4492-bff6-c37b7a09b322 · outbound

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

Improving MLLM Training Efficiency via Stage-Aware Sparsity Flamingo: a visual language model for few-shot learning

Reference 12

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.938540Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation b98092d6-438a-4182-a3d4-5b54c3d21715 · outbound

This paper cites Gpt-4 technical report.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Gpt-4 technical report

Reference 13

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.932009Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 436012f6-a21b-4006-b0d9-1b43f9c99347 · outbound

This paper cites Visual instruction tuning.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Visual instruction tuning

Reference 14

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.935607Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 6623c92a-1315-499e-bacb-3f490b497a14 · outbound

This paper cites Leveraging Visual Tokens for Extended Text Contexts in Multi-Modal Learning.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Leveraging Visual Tokens for Extended Text Contexts in Multi-Modal Learning

Reference 15

Resolution: verified exact; arxiv_id, observed 2026-05-21T22:04:24.222596Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation b849fce7-0d80-48d5-b6ba-cccae4ce6b51 · outbound

This paper cites Qwen2.5-vl technical report.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Qwen2.5-vl technical report

Reference 16

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.910762Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 74f63105-2c8f-4980-90cf-be861f56596a · outbound

This paper cites An image is worth 1/2 tokens after layer 2: Plug-and-play inference acceleration for large vision-language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity An image is worth 1/2 tokens after layer 2: Plug-and-play inference acceleration for large vision-language models

Reference 17

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.916514Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 024760ff-0dcc-4a83-a78d-7002f377df43 · outbound

This paper cites Feather the throttle: Revisiting visual token pruning for vision- language model acceleration.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Feather the throttle: Revisiting visual token pruning for vision- language model acceleration

Reference 18

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.910041Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 3969b073-fcd6-477c-943a-87d3af17b17d · outbound

This paper cites Sparse- vlm: Visual token sparsification for efficient vision-language model inference.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Sparse- vlm: Visual token sparsification for efficient vision-language model inference

Reference 19

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.922118Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 12493ba0-039a-4f87-81c4-66f18fa2e7c8 · outbound

This paper cites Advancing multimodal large language mod- els with quantization-aware scale learning for efficient adap- tation.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Advancing multimodal large language mod- els with quantization-aware scale learning for efficient adap- tation

Reference 20

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.116510Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation ec4881ea-e242-428d-8d87-bb21e00fe79a · outbound

This paper cites Vlmq: Efficient post-training quantization for large vision- language models via hessian augmentation.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Vlmq: Efficient post-training quantization for large vision- language models via hessian augmentation

Reference 21

Resolution: verified exact; arxiv_id, observed 2026-05-21T22:04:24.206029Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 80c4004d-dd03-469b-a046-ef227dd2389b · outbound

This paper cites A multi- level framework for accelerating training transformer models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity A multi- level framework for accelerating training transformer models

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.903743Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 6673568a-47f6-42a4-bda5-5635c99e9e41 · outbound

This paper cites Staged training for transformer lan- guage models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Staged training for transformer lan- guage models

Reference 23

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.127765Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation e7d1e826-9a5d-443d-898c-72b9629fd743 · outbound

This paper cites No train no gain: Revisiting efficient training algorithms for transformer-based language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity No train no gain: Revisiting efficient training algorithms for transformer-based language models

Reference 24

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.121956Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation a941bdcf-8578-40ad-a5a8-180801b9f1cd · outbound

This paper cites Efficient large multi-modal mod- els via visual context compression.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Efficient large multi-modal mod- els via visual context compression

Reference 25

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.119162Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation d13c2368-2272-4a64-a59e-bb5e555a47ae · outbound

This paper cites Pyramiddrop: Accelerating your large vision-language models via pyramid visual redundancy reduction.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Pyramiddrop: Accelerating your large vision-language models via pyramid visual redundancy reduction

Reference 26

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.130526Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation d7aa18bb-d07a-4a15-9bb5-86b45795bc4d · outbound

This paper cites Exploring activation patterns of parameters in lan- guage models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Exploring activation patterns of parameters in lan- guage models

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.939089Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 348f31bb-3b9b-4ff4-be0d-4368203838bf · outbound

This paper cites Mipha: A comprehensive overhaul of multimodal as- sistant with small language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Mipha: A comprehensive overhaul of multimodal as- sistant with small language models

Reference 28

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.941975Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 14c3fd0a-a23c-4b27-a32d-41866300da52 · outbound

This paper cites Moe-llava: Mixture of experts for large vision-language mod- els.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Moe-llava: Mixture of experts for large vision-language mod- els

Reference 29

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.945204Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 076b03f8-c546-402b-b2f6-e83f968bf3d8 · outbound

This paper cites Gqa: A new dataset for real-world visual reasoning and compositional question answering.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Gqa: A new dataset for real-world visual reasoning and compositional question answering

Reference 30

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.948366Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 42c9cbd5-71e3-465b-abd2-37d4eb8ef881 · outbound

This paper cites Making the V in VQA matter: Elevating the role of image understanding in Visual Question Answer- ing.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Making the V in VQA matter: Elevating the role of image understanding in Visual Question Answer- ing

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:05:42.124650Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 25940f83-fed7-47d3-a2a8-beaa8f68caf9 · outbound

This paper cites Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering

Reference 32

Resolution: verified exact; arxiv_id, observed 2026-05-21T22:04:24.213919Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 880ee0b3-6a84-4808-b96d-18099edfb21a · outbound

This paper cites MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models

Reference 33

Resolution: verified exact; local_arxiv, observed 2026-05-21T22:04:24.210033Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Observation 08f7686e-d1fa-48d7-9c9e-094b3e9cec27 · outbound

This paper cites Evaluating object hallucination in large vision-language models.

Improving MLLM Training Efficiency via Stage-Aware Sparsity Evaluating object hallucination in large vision-language models

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-05-21T22:04:24.919608Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-11T06:34:44.6726+00:00.

Pith citing papers

Observation f9bba53d-bd7d-4c60-861f-b3b74aaf6722 · inbound