SharedRep-RLHF: A Shared Representation Approach to RLHF with Diverse Preferences

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Mukherjee, Arpan, Bullo, Marcello, Gündüz, Deniz
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

A Shared Low-Rank Adaptation Approach to Personalized RLHF
por: Liu, Renpu, et al.
Publicado: (2025)

MaxMin-RLHF: Alignment with Diverse Human Preferences
por: Chakraborty, Souradip, et al.
Publicado: (2024)

Sustainable Edge Intelligence Through Energy-Aware Early Exiting
por: Bullo, Marcello, et al.
Publicado: (2023)

Test-time Verification via Optimal Transport: Coverage, ROC, & Sub-optimality
por: Mukherjee, Arpan, et al.
Publicado: (2025)

Offline Constrained RLHF with Multiple Preference Oracles
por: Latham, Brenden, et al.
Publicado: (2026)

Energy-Aware Dynamic Neural Inference
por: Bullo, Marcello, et al.
Publicado: (2024)

Policy-labeled Preference Learning: Is Preference Enough for RLHF?
por: Cho, Taehyun, et al.
Publicado: (2025)

Adaptive Margin RLHF via Preference over Preferences
por: Chittepu, Yaswanth, et al.
Publicado: (2025)

Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF
por: Cen, Shicong, et al.
Publicado: (2024)

Active Preference Optimization for Sample Efficient RLHF
por: Das, Nirjhar, et al.
Publicado: (2024)

Regularized Online RLHF with Generalized Bilinear Preferences
por: Lee, Junghyun, et al.
Publicado: (2026)

WPO: Enhancing RLHF with Weighted Preference Optimization
por: Zhou, Wenxuan, et al.
Publicado: (2024)

Group Robust Preference Optimization in Reward-free RLHF
por: Ramesh, Shyam Sundhar, et al.
Publicado: (2024)

Online Bandit Learning with Offline Preference Data for Improved RLHF
por: Agnihotri, Akhil, et al.
Publicado: (2024)

Influencing Humans to Conform to Preference Models for RLHF
por: Hatgis-Kessell, Stephane, et al.
Publicado: (2025)

RLHF Workflow: From Reward Modeling to Online RLHF
por: Dong, Hanze, et al.
Publicado: (2024)

Understanding the Effects of RLHF on LLM Generalisation and Diversity
por: Kirk, Robert, et al.
Publicado: (2023)

Sharpe Ratio-Guided Active Learning for Preference Optimization in RLHF
por: Belakaria, Syrine, et al.
Publicado: (2025)

RLHF from Heterogeneous Feedback via Personalization and Preference Aggregation
por: Park, Chanwoo, et al.
Publicado: (2024)

Distributional Preference Learning: Understanding and Accounting for Hidden Context in RLHF
por: Siththaranjan, Anand, et al.
Publicado: (2023)

Democratic Preference Alignment via Sortition-Weighted RLHF
por: Sana, Suvadip, et al.
Publicado: (2026)

Understanding the Performance Gap in Preference Learning: A Dichotomy of RLHF and DPO
por: Shi, Ruizhe, et al.
Publicado: (2025)

Factored Causal Representation Learning for Robust Reward Modeling in RLHF
por: Yang, Yupei, et al.
Publicado: (2026)

Mitigating the Alignment Tax of RLHF
por: Lin, Yong, et al.
Publicado: (2023)

OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Framework
por: Hu, Jian, et al.
Publicado: (2024)

Towards Federated RLHF with Aggregated Client Preference for LLMs
por: Wu, Feijie, et al.
Publicado: (2024)

APPA: Adaptive Preference Pluralistic Alignment for Fair Federated RLHF of LLMs
por: Srewa, Mahmoud, et al.
Publicado: (2026)

ROCM: RLHF on consistency models
por: Shekhar, Shivanshu, et al.
Publicado: (2025)

Enhancing RLHF with Human Gaze Modeling
por: Galliamov, Karim, et al.
Publicado: (2025)

Optimal Design for Reward Modeling in RLHF
por: Scheid, Antoine, et al.
Publicado: (2024)

Best-of-Venom: Attacking RLHF by Injecting Poisoned Preference Data
por: Baumgärtner, Tim, et al.
Publicado: (2024)

Towards Data-Centric RLHF: Simple Metrics for Preference Dataset Comparison
por: Shen, Judy Hanwen, et al.
Publicado: (2024)

Advancing Translation Preference Modeling with RLHF: A Step Towards Cost-Effective Solution
por: Xu, Nuo, et al.
Publicado: (2024)

On the Algorithmic Bias of Aligning Large Language Models with RLHF: Preference Collapse and Matching Regularization
por: Xiao, Jiancong, et al.
Publicado: (2024)

Investigating on RLHF methodology
por: Kutalev, Alexey, et al.
Publicado: (2024)

RLHF and IIA: Perverse Incentives
por: Xu, Wanqiao, et al.
Publicado: (2023)

Reward-Robust RLHF in LLMs
por: Yan, Yuzi, et al.
Publicado: (2024)

FedRLHF: A Convergence-Guaranteed Federated Framework for Privacy-Preserving and Personalized RLHF
por: Fan, Flint Xiaofeng, et al.
Publicado: (2024)

Learning a Pessimistic Reward Model in RLHF
por: Xu, Yinglun, et al.
Publicado: (2025)

On a Connection Between Imitation Learning and RLHF
por: Xiao, Teng, et al.
Publicado: (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	"fd6b46637e60a3678af459e7da1c9fd7"
language	"es"
ui	"standard"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790990031.7091 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "fd6b46637e60a3678af459e7da1c9fd7" ]`]
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=fd6b46637e60a3678af459e7da1c9fd7; language=es; 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	"29691"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2509_03672/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2509_03672/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790990031.7048
REQUEST_TIME	1790990031

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler