Learning Like Humans: Advancing LLM Reasoning Capabilities via Adaptive Difficulty Curriculum Learning and Expert-Guided Self-Reformulation

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Zhang, Enci, Yan, Xingang, Lin, Wei, Zhang, Tianxiang, Lu, Qianchun
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Harder Is Better: Boosting Mathematical Reasoning via Difficulty-Aware GRPO and Multi-Aspect Question Reformulation
di: Dai, Yanqi, et al.
Pubblicazione: (2026)

Enhancing LLM Reasoning via Non-Human-Like Reasoning Path Preference Optimization
di: Lu, Junjie, et al.
Pubblicazione: (2025)

MathMixup: Boosting LLM Mathematical Reasoning with Difficulty-Controllable Data Synthesis and Curriculum Learning
di: Li, Xuchen, et al.
Pubblicazione: (2026)

Self-Evolving Curriculum for LLM Reasoning
di: Chen, Xiaoyin, et al.
Pubblicazione: (2025)

CLPO: Curriculum Learning meets Policy Optimization for LLM Reasoning
di: Zhang, Shijie, et al.
Pubblicazione: (2025)

Simulating Human-Like Learning Dynamics with LLM-Empowered Agents
di: Yuan, Yu, et al.
Pubblicazione: (2025)

Rethinking the Sampling Criteria in Reinforcement Learning for LLM Reasoning: A Competence-Difficulty Alignment Perspective
di: Kong, Deyang, et al.
Pubblicazione: (2025)

Know When to Explore: Difficulty-Aware Certainty as a Guide for LLM Reinforcement Learning
di: Li, Ang, et al.
Pubblicazione: (2025)

GanitLLM: Difficulty-Aware Bengali Mathematical Reasoning through Curriculum-GRPO
di: Dipta, Shubhashis Roy, et al.
Pubblicazione: (2026)

DAST: Difficulty-Adaptive Slow-Thinking for Large Reasoning Models
di: Shen, Yi, et al.
Pubblicazione: (2025)

MoL-RL: Distilling Multi-Step Environmental Feedback into LLMs for Feedback-Independent Reasoning
di: Yang, Kang, et al.
Pubblicazione: (2025)

Automatic Curriculum Expert Iteration for Reliable LLM Reasoning
di: Zhao, Zirui, et al.
Pubblicazione: (2024)

Requesting Expert Reasoning: Augmenting LLM Agents with Learned Collaborative Intervention
di: Wang, Zhiming, et al.
Pubblicazione: (2026)

MorphoBench: A Benchmark with Difficulty Adaptive to Model Reasoning
di: Wang, Xukai, et al.
Pubblicazione: (2025)

Learn to Think: Bootstrapping LLM Reasoning Capability Through Graph Representation Learning
di: Gao, Hang, et al.
Pubblicazione: (2025)

LLMCL-GEC: Advancing Grammatical Error Correction with LLM-Driven Curriculum Learning
di: Fang, Tao, et al.
Pubblicazione: (2024)

Make Every Penny Count: Difficulty-Adaptive Self-Consistency for Cost-Efficient Reasoning
di: Wang, Xinglin, et al.
Pubblicazione: (2024)

Capability-aware Prompt Reformulation Learning for Text-to-Image Generation
di: Zhan, Jingtao, et al.
Pubblicazione: (2024)

Mitigating Overthinking in Large Reasoning Models via Difficulty-aware Reinforcement Learning
di: Wan, Qian, et al.
Pubblicazione: (2026)

VL-Cogito: Progressive Curriculum Reinforcement Learning for Advanced Multimodal Reasoning
di: Yuan, Ruifeng, et al.
Pubblicazione: (2025)

DARE: Difficulty-Adaptive Reinforcement Learning with Co-Evolved Difficulty Estimation
di: Zhou, Yang, et al.
Pubblicazione: (2026)

Cog-DRIFT: Exploration on Adaptively Reformulated Instances Enables Learning from Hard Reasoning Problems
di: Chen, Justin Chih-Yao, et al.
Pubblicazione: (2026)

Learning with Challenges: Adaptive Difficulty-Aware Data Generation for Mobile GUI Agent Training
di: Kang, Linjia, et al.
Pubblicazione: (2026)

DART: Difficulty-Adaptive Reasoning Truncation for Efficient Large Language Models
di: Zhang, Ruofan, et al.
Pubblicazione: (2025)

SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution
di: Wei, Yuxiang, et al.
Pubblicazione: (2025)

Discovering Novel LLM Experts via Task-Capability Coevolution
di: Dai, Andrew, et al.
Pubblicazione: (2026)

Tailored Teaching with Balanced Difficulty: Elevating Reasoning in Multimodal Chain-of-Thought via Prompt Curriculum
di: Yang, Xinglong, et al.
Pubblicazione: (2025)

Self-Guided Defense: Adaptive Safety Alignment for Reasoning Models via Synthesized Guidelines
di: Wang, Yuhang, et al.
Pubblicazione: (2025)

EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving
di: Dou, Shihan, et al.
Pubblicazione: (2025)

Graph-R1: Incentivizing the Zero-Shot Graph Learning Capability in LLMs via Explicit Reasoning
di: Wu, Yicong, et al.
Pubblicazione: (2025)

Unlocking Reasoning Capabilities in LLMs via Reinforcement Learning Exploration
di: Deng, Wenhao, et al.
Pubblicazione: (2025)

Curriculum Reinforcement Learning from Easy to Hard Tasks Improves LLM Reasoning
di: Parashar, Shubham, et al.
Pubblicazione: (2025)

From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning
di: Jiang, Xitai, et al.
Pubblicazione: (2026)

The LLM Already Knows: Estimating LLM-Perceived Question Difficulty via Hidden Representations
di: Zhu, Yubo, et al.
Pubblicazione: (2025)

Adaptive Multi-Expert Reasoning via Difficulty-Aware Routing and Uncertainty-Guided Aggregation
di: Ehab, Mohamed, et al.
Pubblicazione: (2026)

VAPO: Efficient and Reliable Reinforcement Learning for Advanced Reasoning Tasks
di: Yue, Yu, et al.
Pubblicazione: (2025)

StoryCoder: Narrative Reformulation for Structured Reasoning in LLM Code Generation
di: Jang, Geonhui, et al.
Pubblicazione: (2026)

Human-Aware Robot Navigation via Reinforcement Learning with Hindsight Experience Replay and Curriculum Learning
di: Li, Keyu, et al.
Pubblicazione: (2021)

WorkForceAgent-R1: Incentivizing Reasoning Capability in LLM-based Web Agents via Reinforcement Learning
di: Zhuang, Yuchen, et al.
Pubblicazione: (2025)

Online Difficulty Filtering for Reasoning Oriented Reinforcement Learning
di: Bae, Sanghwan, 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
ui	"standard"
VUFIND_SESSION	"ed76789417e352e3e983bb8391524723"
language	"it"

Session

Key	Value
__Laminas	array:2 [`"_REQUEST_ACCESS_TIME" => 1790867837.9583 "_VALID" => array:1 [ "Laminas\Session\Validator\Id" => "ed76789417e352e3e983bb8391524723" ]`]
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	"ui=standard; VUFIND_SESSION=ed76789417e352e3e983bb8391524723; language=it"
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	"216.73.216.147"
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	"28841"
REDIRECT_URL	""" /Record/\n arxiv_https___arxiv_org_abs_2505_08364/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_08364/Similar"
SCRIPT_NAME	"/index.php"
PHP_SELF	"/index.php"
REQUEST_TIME_FLOAT	1790867837.9541
REQUEST_TIME	1790867837

Environment Variables empty

Registered Handlers

0. Whoops\Handler\PrettyPageHandler