M3PO: Multimodal-Model-Guided Preference Optimization for Visual Instruction Following

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Gao, Ruirui, Johnson, Emily, Tan, Bowen, Qian, Yanfei
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Computation and Language
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Reverse Preference Optimization for Complex Instruction Following
di: Huang, Xiang, et al.
Pubblicazione: (2025)

PrefPO: Pairwise Preference Prompt Optimization
di: Singhal, Rahul, et al.
Pubblicazione: (2026)

ConfPO: Exploiting Policy Model Confidence for Critical Token Selection in Preference Optimization
di: Yoon, Hee Suk, et al.
Pubblicazione: (2025)

AMaPO: Adaptive Margin-attached Preference Optimization for Language Model Alignment
di: Deng, Ruibo, et al.
Pubblicazione: (2025)

MMedPO: Aligning Medical Vision-Language Models with Clinical-Aware Multimodal Preference Optimization
di: Zhu, Kangyu, et al.
Pubblicazione: (2024)

FocalPO: Enhancing Preference Optimizing by Focusing on Correct Preference Rankings
di: Liu, Tong, et al.
Pubblicazione: (2025)

SeaPO: Strategic Error Amplification for Robust Preference Optimization of Large Language Models
di: Rao, Jun, et al.
Pubblicazione: (2025)

CiPO: Counterfactual Unlearning for Large Reasoning Models through Iterative Preference Optimization
di: Li, Junyi, et al.
Pubblicazione: (2026)

IOPO: Empowering LLMs with Complex Instruction Following via Input-Output Preference Optimization
di: Zhang, Xinghua, et al.
Pubblicazione: (2024)

ComPO: Community Preferences for Language Model Personalization
di: Kumar, Sachin, et al.
Pubblicazione: (2024)

MIA-Bench: Towards Better Instruction Following Evaluation of Multimodal LLMs
di: Qian, Yusu, et al.
Pubblicazione: (2024)

IHEval: Evaluating Language Models on Following the Instruction Hierarchy
di: Zhang, Zhihan, et al.
Pubblicazione: (2025)

A Systematic Examination of Preference Learning through the Lens of Instruction-Following
di: Kim, Joongwon, et al.
Pubblicazione: (2024)

$f$-PO: Generalizing Preference Optimization with $f$-divergence Minimization
di: Han, Jiaqi, et al.
Pubblicazione: (2024)

LiPO: Listwise Preference Optimization through Learning-to-Rank
di: Liu, Tianqi, et al.
Pubblicazione: (2024)

Reasoning Strategies in Large Language Models: Can They Follow, Prefer, and Optimize?
di: Zhang, Yanjian, et al.
Pubblicazione: (2025)

DiffPO: Diffusion-styled Preference Optimization for Efficient Inference-Time Alignment of Large Language Models
di: Chen, Ruizhe, et al.
Pubblicazione: (2025)

Improving the Accuracy and Efficiency of Legal Document Tagging with Large Language Models and Instruction Prompts
di: Johnson, Emily, et al.
Pubblicazione: (2025)

DecIF: Improving Instruction-Following through Meta-Decomposition
di: Hui, Tingfeng, et al.
Pubblicazione: (2025)

Text as Images: Can Multimodal Large Language Models Follow Printed Instructions in Pixels?
di: Li, Xiujun, et al.
Pubblicazione: (2023)

SimPO: Simple Preference Optimization with a Reference-Free Reward
di: Meng, Yu, et al.
Pubblicazione: (2024)

HREF: Human Response-Guided Evaluation of Instruction Following in Language Models
di: Lyu, Xinxi, et al.
Pubblicazione: (2024)

PerPO: Perceptual Preference Optimization via Discriminative Rewarding
di: Zhu, Zining, et al.
Pubblicazione: (2025)

Sparkles: Unlocking Chats Across Multiple Images for Multimodal Instruction-Following Models
di: Huang, Yupan, et al.
Pubblicazione: (2023)

MidPO: Dual Preference Optimization for Safety and Helpfulness in Large Language Models via a Mixture of Experts Framework
di: Qi, Yupeng, et al.
Pubblicazione: (2025)

RankPO: Preference Optimization for Job-Talent Matching
di: Zhang, Yafei, et al.
Pubblicazione: (2025)

Balancing Continuous Pre-Training and Instruction Fine-Tuning: Optimizing Instruction-Following in LLMs
di: Jindal, Ishan, et al.
Pubblicazione: (2024)

CoGenesis: A Framework Collaborating Large and Small Language Models for Secure Context-Aware Instruction Following
di: Zhang, Kaiyan, et al.
Pubblicazione: (2024)

Self-Steering Optimization: Autonomous Preference Optimization for Large Language Models
di: Xiang, Hao, et al.
Pubblicazione: (2024)

AIR: A Systematic Analysis of Annotations, Instructions, and Response Pairs in Preference Dataset
di: He, Bingxiang, et al.
Pubblicazione: (2025)

SoLoPO: Unlocking Long-Context Capabilities in LLMs via Short-to-Long Preference Optimization
di: Sun, Huashan, et al.
Pubblicazione: (2025)

KnowPO: Knowledge-aware Preference Optimization for Controllable Knowledge Selection in Retrieval-Augmented Language Models
di: Zhang, Ruizhe, et al.
Pubblicazione: (2024)

Evaluating Large Language Models at Evaluating Instruction Following
di: Zeng, Zhiyuan, et al.
Pubblicazione: (2023)

Pragmatic Instruction Following and Goal Assistance via Cooperative Language-Guided Inverse Planning
di: Zhi-Xuan, Tan, et al.
Pubblicazione: (2024)

Instruction Following without Instruction Tuning
di: Hewitt, John, et al.
Pubblicazione: (2024)

Automatic Layout Planning for Visually-Rich Documents with Instruction-Following Models
di: Zhu, Wanrong, et al.
Pubblicazione: (2024)

TAB-PO: Preference Optimization with a Token-Level Adaptive Barrier for Token-Critical Structured Generation
di: Fodeh, Samah, et al.
Pubblicazione: (2026)

Advantage-Guided Distillation for Preference Alignment in Small Language Models
di: Gao, Shiping, et al.
Pubblicazione: (2025)

MDCure: A Scalable Pipeline for Multi-Document Instruction-Following
di: Liu, Gabrielle Kaili-May, et al.
Pubblicazione: (2024)

M-IFEval: Multilingual Instruction-Following Evaluation
di: Dussolle, Antoine, et al.
Pubblicazione: (2025)

if (filemtime($filename) + $maxlifetime < time()) { unlink($filename); $count++; } } return $count; } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ protected function saveSession($sessId, $data): bool { $sessFile = $this->path . '/sess_' . $sessId; if ($handle = fopen($sessFile, 'w')) { $return = false; // Lock the file for exclusive access to avoid issues with multiple // processes writing session simultaneously: if (flock($handle, LOCK_EX)) { $return = fwrite($handle, $data); // Make sure that there's no trailing data by truncating the file to // the correct length: ftruncate($handle, strlen($data)); fflush($handle); flock($handle, LOCK_UN); } fclose($handle); if ($return !== false) { return true; } } // If we got this far, something went wrong with the file output... // It is tempting to throw an exception here, but this code is called // outside of the context of exception handling, so all we can do is // echo a message.

// Anecdotal testing Today and Yesterday seems to indicate destroy() // is called by the garbage collector and everything is good. // Something to keep in mind though. return 0; } /** * Write function that is called when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ public function write($sessId, $data): bool { if ($this->writesDisabled) { return true; } return $this->saveSession($sessId, $data); } /** * A function that is called internally when session data is to be saved. * * @param string $sessId The current session ID * @param string $data The session data to write * * @return bool */ abstract protected function saveSession($sessId, $data): bool; }

* * @return void */ public function writeClose() { // The assumption is that we're using PHP's ext/session. // session_write_close() will actually overwrite $_SESSION with an // empty array on completion -- which leads to a mismatch between what // is in the storage object and $_SESSION. To get around this, we // temporarily reset $_SESSION to an array, and then re-link it to // the storage object. // // Additionally, while you _can_ write to $_SESSION following a // session_write_close() operation, no changes made to it will be // flushed to the session handler. As such, we now mark the storage // object isImmutable. $storage = $this->getStorage(); if (! $storage->isImmutable()) { $_SESSION = $storage->toArray(true); session_write_close(); $storage->fromArray($_SESSION); $storage->markImmutable(); } } /** * Attempt to set the session name * * If the session has already been started, or if the name provided fails * validation, an exception will be raised. * * @param string $name * @return SessionManager * @throws Exception\InvalidArgumentException */ public function setName($name) { if ($this->sessionExists()) { throw new Exception\InvalidArgumentException( 'Cannot set session name after a session has already started'

Environment & details:

GET Data empty

POST Data empty

Files empty

Key	Value
VUFIND_SESSION	"df2d858ae13d18420aaa35382b8b8c4b"
language	"it"
ui	"standard"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790990194.2214 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "df2d858ae13d18420aaa35382b8b8c4b" ]`]
SessionState	Laminas\Stdlib\ArrayObject {#938}
FlashMessenger	Laminas\Stdlib\ArrayObject {#1433}
SessionHelper	Laminas\Stdlib\ArrayObject {#1759}

Server/Request Data

Key	Value
REDIRECT_VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
REDIRECT_HTTPS	"on"
REDIRECT_SSL_TLS_SNI	"info.umar.mx"
REDIRECT_STATUS	"200"
VUFIND_LOCAL_DIR	"/usr/local/vufind/local"
HTTPS	"on"
SSL_TLS_SNI	"info.umar.mx"
HTTP_ACCEPT	"/"
HTTP_USER_AGENT	"Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)"
HTTP_COOKIE	"VUFIND_SESSION=df2d858ae13d18420aaa35382b8b8c4b; language=it; ui=standard"
HTTP_ACCEPT_ENCODING	"gzip, br, zstd, deflate"
HTTP_HOST	"info.umar.mx"
PATH	"/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/snap/bin"
SERVER_SIGNATURE	"<address>Apache/2.4.66 (Ubuntu) Server at info.umar.mx Port 443</address>\n"
SERVER_SOFTWARE	"Apache/2.4.66 (Ubuntu)"
SERVER_NAME	"info.umar.mx"
SERVER_ADDR	"200.23.222.19"
SERVER_PORT	"443"
REMOTE_ADDR	"216.73.217.142"
DOCUMENT_ROOT	"/usr/local/vufind/public"
REQUEST_SCHEME	"https"
CONTEXT_PREFIX	""
CONTEXT_DOCUMENT_ROOT	"/usr/local/vufind/public"
SERVER_ADMIN	"admin@localhost"
SCRIPT_FILENAME	"/usr/local/vufind/public/index.php"
REMOTE_PORT	"50522"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2508_12458/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2508_12458/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790990194.2142
REQUEST_TIME	1790990194

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler