Cover Image

From $f(x)$ and $g(x)$ to $f(g(x))$: LLMs Learn New Skills in RL by Composing Old Ones

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Yuan, Lifan, Chen, Weize, Zhang, Yuchen, Cui, Ganqu, Wang, Hanbin, You, Ziming, Ding, Ning, Liu, Zhiyuan, Sun, Maosong, Peng, Hao
Format:	Preprint
Published:	2025
Subjects:	Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

On the monodromy action for $f(x,y)=g(x)+h(y)$
by: Garcia, Daniel López, et al.
Published: (2023)

Faulhaber's formula, Bernoulli numbers, power sums of natural numbers and totatives and the functional equation $f(x)+x^k=f(x+1)$
by: Wu, Chai Wah
Published: (2024)

Both orbital and spin torques originate from r x g
by: Crimin, Frances, et al.
Published: (2024)

A characterization on $(g,f)$-parity orientations
by: Lu, Hongliang, et al.
Published: (2024)

JustRL: Scaling a 1.5B LLM with a Simple RL Recipe
by: He, Bingxiang, et al.
Published: (2025)

A Proof of Exact Convergence Rate of Gradient Descent. Part I. Performance Criterion $\Vert \nabla f(x_N)\Vert^2/(f(x_0)-f_*)$
by: Kim, Jungbin
Published: (2024)

Advancing LLM Reasoning Generalists with Preference Trees
by: Yuan, Lifan, et al.
Published: (2024)

INTERVENOR: Prompting the Coding Ability of Large Language Models with the Interactive Chain of Repair
by: Wang, Hanbin, et al.
Published: (2023)

Free Process Rewards without Process Labels
by: Yuan, Lifan, et al.
Published: (2024)

On existence questions for the functional equations $f^8+g^8+h^8=1$ and $f^6+g^6+h^6=1$
by: Li, Xiao-Min, et al.
Published: (2024)

Anomalous 4$f$ fine structure in TmSe$_{1-x}$Te$_x$ across the metal-insulator transition
by: Min, C. -H., et al.
Published: (2024)

On the singular elliptic problem involving the variable order fractional Musielak $g_{x,y}$-Laplacian
by: Baalal, Azeddine, et al.
Published: (2025)

The three-loop single-mass heavy-flavor corrections to the structure functions $F_2(x,Q^2)$ and $g_1(x,Q^2)$
by: Ablinger, J., et al.
Published: (2025)

The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models
by: Cui, Ganqu, et al.
Published: (2025)

New Words or Redefined Old Ones?
by: Sven Ove Hansson
Published: (2026)

Order of the SU(N_f) x SU(N_f) chiral transition via the functional renormalization group
by: Fejos, G., et al.
Published: (2024)

The Davenport-Lewis-Schinzel problem on the reducibility of $f(X)-g(Y)$
by: Behajaina, Angelot, et al.
Published: (2026)

Al 1‐x Sc x N‐Based Ferroelectric Domain‐Wall Memristors
by: Haidong Lu, et al.
Published: (2025)

UltraFeedback: Boosting Language Models with Scaled AI Feedback
by: Cui, Ganqu, et al.
Published: (2023)

Controllable Preference Optimization: Toward Controllable Multi-Objective Alignment
by: Guo, Yiju, et al.
Published: (2024)

First‐Principles Study of Structural, Elastic, and Piezoelectric Properties of Wurtzite Superlattice Sc x Al 1− x N, Sc x Ga 1− x N, and Sc x In 1− x N Alloys
by: Hang Cui, et al.
Published: (2026)

$x(1-t(x+x^{-1}))F(x;t) = x-tF(0;t)$
by: Buchacher, Manfred
Published: (2025)

Process Reinforcement through Implicit Rewards
by: Cui, Ganqu, et al.
Published: (2025)

Temperature and Electron Concentration Dependences of 1/f Noise in Hg$_{1-x}$Cd$_x$Te -- Evidence for a Mobility Fluctuations Mechanism
by: Rehman, Adil, et al.
Published: (2024)

RLPR: Extrapolating RLVR to General Domains without Verifiers
by: Yu, Tianyu, et al.
Published: (2025)

Rings for Which f.g. Projective Modules Have the FI-extending Property
by: Danchev, Peter, et al.
Published: (2025)

Mastering Text, Code and Math Simultaneously via Fusing Highly Specialized Language Models
by: Ding, Ning, et al.
Published: (2024)

The Overthinker's DIET: Cutting Token Calories with DIfficulty-AwarE Training
by: Chen, Weize, et al.
Published: (2025)

Analytic Solution for the Helicity Evolution Equations at Small $x$ and Large $N_c\&N_f$
by: Borden, Jeremy, et al.
Published: (2025)

The f–x domain Transformer network for multi‐frequency ground‐penetrating radar data fusion
by: Xuebing Zhang, et al.
Published: (2026)

Quantum geometry and $X$-wave magnets with $X=p,d,f,g,i$
by: Ezawa, Motohiko
Published: (2025)

Graded turbidite record of ICDP core 5017-1, (Fig. 3f-g)
by: Lu, Yin, et al.
Published: (2021)

The Right Time Matters: Data Arrangement Affects Zero-Shot Generalization in Instruction Tuning
by: He, Bingxiang, et al.
Published: (2024)

Global Existence and Finite-Time Blow-Up of Solutions for Parabolic Equations Involving the Fractional Musielak $g_{x,y}$-Laplacian
by: Arora, Rakesh, et al.
Published: (2025)

On the linear (in)dependence of sequences of derivatives of the functions $x^n\sin x$ and $x^n\cos x$
by: Fecenko, Jozef, et al.
Published: (2023)

On the stability of $\ddot x(t)+α(t)\dot x(t)+β(t) x(t)=0$
by: Bessa, Mario, et al.
Published: (2024)

On powers of the diophantine function $\star:x\mapsto x(x+1)$
by: Silberger, Donald
Published: (2024)

Effects of oxygen vacancy on bond ionicity and microwave dielectric properties of Y x Ce 1− x O 2−0.5 x ( x = 0–0.7) ceramics
by: Ming‐An Shi, et al.
Published: (2024)

On the sign changes of $ψ(x)-x$
by: Grześkowiak, Maciej, et al.
Published: (2024)

Off-Diagonal Continuous Rado Numbers $x_1 + x_2 + \dots + x_k = x_0$
by: Vestal, Don, et al.
Published: (2025)