Learning to Plan & Reason for Evaluation with Thinking-LLM-as-a-Judge

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Saha, Swarnadeep, Li, Xian, Ghazvininejad, Marjan, Weston, Jason, Wang, Tianlu
Format:	Preprint
Publié:	2025
Sujets:	Artificial Intelligence 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

J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learning
par: Whitehouse, Chenxi, et autres
Publié: (2025)

Branch-Solve-Merge Improves Large Language Model Evaluation and Generation
par: Saha, Swarnadeep, et autres
Publié: (2023)

Contextual Position Encoding: Learning to Count What's Important
par: Golovneva, Olga, et autres
Publié: (2024)

Reasoning over mathematical objects: on-policy reward modeling and test time aggregation
par: Aggarwal, Pranjal, et autres
Publié: (2026)

Think-J: Learning to Think for Generative LLM-as-a-Judge
par: Huang, Hui, et autres
Publié: (2025)

StepWiser: Stepwise Generative Judges for Wiser Reasoning
par: Xiong, Wei, et autres
Publié: (2025)

Self-Taught Evaluators
par: Wang, Tianlu, et autres
Publié: (2024)

ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs
par: Chen, Justin Chih-Yao, et autres
Publié: (2023)

Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge
par: Wu, Tianhao, et autres
Publié: (2024)

LeMAJ (Legal LLM-as-a-Judge): Bridging Legal Reasoning and LLM Evaluation
par: Enguehard, Joseph, et autres
Publié: (2025)

Thinking Fast, Thinking Wrong: Intuitiveness Modulates LLM Counterfactual Reasoning in Policy Evaluation
par: He, Yanjie
Publié: (2026)

Multimodal RewardBench: Holistic Evaluation of Reward Models for Vision Language Models
par: Yasunaga, Michihiro, et autres
Publié: (2025)

System-1.x: Learning to Balance Fast and Slow Planning with Language Models
par: Saha, Swarnadeep, et autres
Publié: (2024)

Thinking LLMs: General Instruction Following with Thought Generation
par: Wu, Tianhao, et autres
Publié: (2024)

OptimalThinkingBench: Evaluating Over and Underthinking in LLMs
par: Aggarwal, Pranjal, et autres
Publié: (2025)

Self-Challenging Language Model Agents
par: Zhou, Yifei, et autres
Publié: (2025)

Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey
par: Zhu, Jason, et autres
Publié: (2025)

CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks
par: Yu, Ping, et autres
Publié: (2025)

Evaluating Metrics for Safety with LLM-as-Judges
par: Clegg, Kester, et autres
Publié: (2025)

Wide-Horizon Thinking and Simulation-Based Evaluation for Real-World LLM Planning with Multifaceted Constraints
par: Yang, Dongjie, et autres
Publié: (2025)

JudgeLRM: Large Reasoning Models as a Judge
par: Chen, Nuo, et autres
Publié: (2025)

JudgeBoard: Benchmarking and Enhancing Small Language Models for Reasoning Evaluation
par: Bi, Zhenyu, et autres
Publié: (2025)

LLM-BABYBENCH: Understanding and Evaluating Grounded Planning and Reasoning in LLMs
par: Choukrani, Omar, et autres
Publié: (2025)

Plan Then Action:High-Level Planning Guidance Reinforcement Learning for LLM Reasoning
par: Dou, Zhihao, et autres
Publié: (2025)

Thinkless: LLM Learns When to Think
par: Fang, Gongfan, et autres
Publié: (2025)

Jointly Reinforcing Diversity and Quality in Language Model Generations
par: Li, Tianjian, et autres
Publié: (2025)

Through the Judge's Eyes: Inferred Thinking Traces Improve Reliability of LLM Raters
par: Zhang, Xingjian, et autres
Publié: (2025)

JudgeBench: A Benchmark for Evaluating LLM-based Judges
par: Tan, Sijun, et autres
Publié: (2024)

TrustJudge: Inconsistencies of LLM-as-a-Judge and How to Alleviate Them
par: Wang, Yidong, et autres
Publié: (2025)

Iterative Reasoning Preference Optimization
par: Pang, Richard Yuanzhe, et autres
Publié: (2024)

Thinking About Thinking: Evaluating Reasoning in Post-Trained Language Models
par: Singla, Pratham, et autres
Publié: (2025)

A Survey on LLM-as-a-Judge
par: Gu, Jiawei, et autres
Publié: (2024)

Counterargument for Critical Thinking as Judged by AI and Humans
par: Adewumi, Tosin, et autres
Publié: (2026)

VERT: Reliable LLM Judges for Radiology Report Evaluation
par: Bologna, Federica, et autres
Publié: (2026)

LLM-RadJudge: Achieving Radiologist-Level Evaluation for X-Ray Report Generation
par: Wang, Zilong, et autres
Publié: (2024)

JudgeAgent: Beyond Static Benchmarks for Knowledge-Driven and Dynamic LLM Evaluation
par: Shi, Zhichao, et autres
Publié: (2025)

Judging the Judges: A Systematic Study of Position Bias in LLM-as-a-Judge
par: Shi, Lin, et autres
Publié: (2024)

Branch-Train-MiX: Mixing Expert LLMs into a Mixture-of-Experts LLM
par: Sukhbaatar, Sainbayar, et autres
Publié: (2024)

System-Level Natural Language Feedback
par: Yuan, Weizhe, et autres
Publié: (2023)

When Metrics Disagree: Automatic Similarity vs. LLM-as-a-Judge for Clinical Dialogue Evaluation
par: Sun, Bian, et autres
Publié: (2026)

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	"8e9c674bfa6c224f50f0861b8975c5e8"
ui	"standard"
language	"fr"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790882332.7632 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "8e9c674bfa6c224f50f0861b8975c5e8" ]`]
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=8e9c674bfa6c224f50f0861b8975c5e8; ui=standard; language=fr"
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	"8389"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2501_18099/Similar """
GATEWAY_INTERFACE	"CGI/1.1"
SERVER_PROTOCOL	"HTTP/1.1"
REQUEST_METHOD	"GET"
QUERY_STRING	""
REQUEST_URI	"/Record/%0Aarxiv_https___arxiv_org_abs_2501_18099/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790882332.7571
REQUEST_TIME	1790882332

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler