Enhancing Efficiency and Exploration in Reinforcement Learning for LLMs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Liao, Mengqi, Xi, Xiangyu, Chen, Ruinian, Leng, Jia, Hu, Yangen, Zeng, Ke, Liu, Shuai, Wan, Huaiyu
Format:	Preprint
Publié:	2025
Sujets:	Machine Learning Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Enhancing the Code Reasoning Capabilities of LLMs via Consistency-based Reinforcement Learning
par: Qin, Zhanyue, et autres
Publié: (2026)

QeRL: Beyond Efficiency -- Quantization-enhanced Reinforcement Learning for LLMs
par: Huang, Wei, et autres
Publié: (2025)

DenoiseRotator: Enhance Pruning Robustness for LLMs via Importance Concentration
par: Gu, Tianteng, et autres
Publié: (2025)

Reinforcement Learning Enhanced LLMs: A Survey
par: Wang, Shuhe, et autres
Publié: (2024)

Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration
par: Deng, Wenhao, et autres
Publié: (2025)

Optimizing Large Language Models with an Enhanced LoRA Fine-Tuning Algorithm for Efficiency and Robustness in NLP Tasks
par: Hu, Jiacheng, et autres
Publié: (2024)

BroRL: Scaling Reinforcement Learning via Broadened Exploration
par: Hu, Jian, et autres
Publié: (2025)

S$^2$R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learning
par: Ma, Ruotian, et autres
Publié: (2025)

Large Language Model Prompt Datasets: An In-depth Analysis and Insights
par: Zhang, Yuanming, et autres
Publié: (2025)

Reasoning through Exploration: A Reinforcement Learning Framework for Robust Function Calling
par: Hao, Bingguang, et autres
Publié: (2025)

Towards Understanding Multi-Task Learning (Generalization) of LLMs via Detecting and Exploring Task-Specific Neurons
par: Leng, Yongqi, et autres
Publié: (2024)

When Bias Pretends to Be Truth: How Spurious Correlations Undermine Hallucination Detection in LLMs
par: Wang, Shaowen, et autres
Publié: (2025)

Reinforcement Learning for Tool-Integrated Interleaved Thinking towards Cross-Domain Generalization
par: Chen, Zhengyu, et autres
Publié: (2025)

AutoTriton: Automatic Triton Programming with Reinforcement Learning in LLMs
par: Li, Shangzhan, et autres
Publié: (2025)

Enhancing LLM Knowledge Learning through Generalization
par: Zhu, Mingkang, et autres
Publié: (2025)

Towards High Data Efficiency in Reinforcement Learning with Verifiable Reward
par: Tang, Xinyu, et autres
Publié: (2025)

Exploration Hacking: Can LLMs Learn to Resist RL Training?
par: Jang, Eyon, et autres
Publié: (2026)

ConfClip: Confidence-Weighted and Clipped Reward for Reinforcement Learning in LLMs
par: Zhang, Bonan, et autres
Publié: (2025)

Stabilizing Reinforcement Learning with LLMs: Formulation and Practices
par: Zheng, Chujie, et autres
Publié: (2025)

Self-Hinting Language Models Enhance Reinforcement Learning
par: Liao, Baohao, et autres
Publié: (2026)

Stratified GRPO: Handling Structural Heterogeneity in Reinforcement Learning of LLM Search Agents
par: Zhu, Mingkang, et autres
Publié: (2025)

Open-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model
par: Hu, Jingcheng, et autres
Publié: (2025)

Low-probability Tokens Sustain Exploration in Reinforcement Learning with Verifiable Reward
par: Huang, Guanhua, et autres
Publié: (2025)

TemplateRL: Structured Template-Guided Reinforcement Learning for LLM Reasoning
par: Wu, Jinyang, et autres
Publié: (2025)

SPS: Steering Probability Squeezing for Better Exploration in Reinforcement Learning for Large Language Models
par: Huo, Yifu, et autres
Publié: (2026)

Improving Sample Efficiency of Reinforcement Learning with Background Knowledge from Large Language Models
par: Zhang, Fuxiang, et autres
Publié: (2024)

TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning
par: Wei, Zhepei, et autres
Publié: (2025)

Free Energy-Driven Reinforcement Learning with Adaptive Advantage Shaping for Unsupervised Reasoning in LLMs
par: Huang, Yiming, et autres
Publié: (2026)

ToolExpander: Extending the Frontiers of Tool-Using Reinforcement Learning to Weak LLMs
par: Chen, Fu, et autres
Publié: (2025)

ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning
par: Wan, Ziyu, et autres
Publié: (2025)

DISPO: Enhancing Training Efficiency and Stability in Reinforcement Learning for Large Language Model Mathematical Reasoning
par: Karaman, Batuhan K., et autres
Publié: (2026)

Back to Basics: Revisiting Exploration in Reinforcement Learning for LLM Reasoning via Generative Probabilities
par: Li, Pengyi, et autres
Publié: (2026)

Token Hidden Reward: Steering Exploration-Exploitation in Group Relative Deep Reinforcement Learning
par: Deng, Wenlong, et autres
Publié: (2025)

CDE: Curiosity-Driven Exploration for Efficient Reinforcement Learning in Large Language Models
par: Dai, Runpeng, et autres
Publié: (2025)

One Sample to Rule Them All: Extreme Data Efficiency in Multidiscipline Reasoning with Reinforcement Learning
par: Li, Yiyuan, et autres
Publié: (2026)

Aligning Frozen LLMs by Reinforcement Learning: An Iterative Reweight-then-Optimize Approach
par: Zhang, Xinnan, et autres
Publié: (2025)

Internalizing Outcome Supervision into Process Supervision: A New Paradigm for Reinforcement Learning for Reasoning
par: Ding, Fei, et autres
Publié: (2026)

Deep Dense Exploration for LLM Reinforcement Learning via Pivot-Driven Resampling
par: Guo, Yiran, et autres
Publié: (2026)

TRN-R1-Zero: Text-rich Network Reasoning via LLMs with Reinforcement Learning Only
par: Liu, Yilun, et autres
Publié: (2026)

Natural Language Reinforcement Learning
par: Feng, Xidong, et autres
Publié: (2024)

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
language	"fr"
VUFIND_SESSION	"c046ea49ecb9f06d05e24a94f8b49925"
ui	"standard"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790987125.9483 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "c046ea49ecb9f06d05e24a94f8b49925" ]`]
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	"language=fr; VUFIND_SESSION=c046ea49ecb9f06d05e24a94f8b49925; 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	"40907"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2505_18573/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2505_18573/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790987125.943
REQUEST_TIME	1790987125

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler