Papers › SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for Multi-modal...

SPHINX: The Joint Mixing of Weights, Tasks, and Visual Embeddings for Multi-modal Large Language Models

13 Nov 2023arXiv:2311.07575archive 2025-07-28

Ziyi Lin, Chris Liu, Renrui Zhang, Peng Gao, Longtian Qiu, Han Xiao, Han Qiu, Chen Lin, Wenqi Shao, Keqin Chen, Jiaming Han, Siyuan Huang, Yichi Zhang, Xuming He, Hongsheng Li, Yu Qiao

We present SPHINX, a versatile multi-modal large language model (MLLM) with a joint mixing of model weights, tuning tasks, and visual embeddings. First, for stronger vision-language alignment, we unfreeze the large language model (LLM) during pre-training, and introduce a weight mix strategy between LLMs trained by real-world and synthetic data. By directly integrating the weights from two domains, the mixed LLM can efficiently incorporate diverse semantics with favorable robustness. Then, to enable multi-purpose capabilities, we mix a variety of tasks for joint visual instruction tuning, and design task-specific instructions to avoid inter-task conflict. In addition to the basic visual question answering, we include more challenging tasks such as region-level understanding, caption grounding, document layout detection, and human pose estimation, contributing to mutual enhancement over different scenarios. Additionally, we propose to extract comprehensive visual embeddings from various network architectures, pre-training paradigms, and information granularity, providing language models with more robust image representations. Based on our proposed joint mixing, SPHINX exhibits superior multi-modal understanding capabilities on a wide range of applications. On top of this, we further propose an efficient strategy aiming to better capture fine-grained appearances of high-resolution images. With a mixing of different scales and high-resolution sub-images, SPHINX attains exceptional visual parsing and reasoning performance on existing evaluation benchmarks. We hope our work may cast a light on the exploration of joint mixing in future MLLM research. Code is released at https://github.com/Alpha-VLLM/LLaMA2-Accessory.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology: its repositories, every harvested function with whether it ran, its licence and the call to fetch it.

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="2311.07575")

Code

Syntology Ran 3 of 3 code samples harvested from 0 repositories linked to this paper; 0 have no recorded run. Of those that ran: 1 ran · violated contract; 1 ran · our draft was wrong; 1 ran · fixture could not drive it.

By repository: 3 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

alpha-vllm/llama2-accessory officialmentioned in paperpytorch 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

3 samples harvested; 3 ran; 0 honoured the contract we drafted; 0 have no recorded run. Read from Syntology's graph 2026-09-25; that is when this build read the record, not when the samples ran.

1ran · violated contract
1ran · our draft was wrong
1ran · fixture could not drive it

Licence: 3 of the 3 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.

Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “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.

apply_rotary_emb identical code first harvested elsewhere ran · our draft was wrong licence of this copy not recorded · dcae39c8928b5fd8 · report
precompute_freqs_cis identical code first harvested elsewhere ran · violated contract fingerprinted licence of this copy not recorded · 2c4423db8989ee05 · report
reshape_for_broadcast identical code first harvested elsewhere ran · fixture could not drive it fingerprinted licence of this copy not recorded · 70bf6ebaafd266c4 · report

Tasks

Described Object DetectionLanguage ModelingLanguage ModellingLarge Language ModelPose EstimationQuestion AnsweringVisual Question AnsweringVisual Question Answering (VQA)

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Described Object Detection Description Detection Dataset SPHINX-7B Intra-scenario ABS mAP 7.9 #6 of 8 Archive leaderboard report
Described Object Detection Description Detection Dataset SPHINX-7B Intra-scenario FULL mAP 10.6 #6 of 8 Archive leaderboard report
Described Object Detection Description Detection Dataset SPHINX-7B Intra-scenario PRES mAP 11.4 #6 of 8 Archive leaderboard report
Visual Question Answering BenchLMM Sphinx-V2-1K GPT-3.5 score 57.43 #2 of 10 Archive leaderboard report
Visual Question Answering MM-Vet SPHINX-2k GPT-4 score 40.2 #109 of 231 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval SPHINX v2 Abductive 49.85 #2 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval SPHINX v2 Analogical 20.69 #2 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval SPHINX v2 Deductive 42.17 #2 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval SPHINX v2 Overall score 39.48 #2 of 14 Archive leaderboard report
Visual Question Answering (VQA) InfiMM-Eval SPHINX v2 Params 16B #2 of 14 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Visual Parsing

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