Trust Region Masking for Long-Horizon LLM Reinforcement Learning

arXiv:2512.23075v4 Announce Type: replace-cross
Abstract: Policy gradient methods for Large Language Models optimize a policy $pi_theta$ via a surrogate objective computed from samples of a rollout policy $pi_{text{roll}}$. However, modern LLM-RL pipelines suffer from unavoidable implementation divergences — backend discrepancies, Mixture-of-Experts routing discontinuities, and distributed training staleness — causing off-policy mismatch ($pi_{text{roll}} neq pi_theta$) and approximation errors between the surrogate and the true objective. We demonstrate that classical trust region bounds on this error scale as $O(T^2)$ with sequence length $T$, rendering them vacuous for long-horizon tasks. To address this, we derive a family of bounds — both KL-based and TV-based — including a Pinsker-Marginal bound ($O(T^{3/2})$), a Mixed bound ($O(T)$), and an Adaptive bound that strictly generalizes the Pinsker-Marginal bound via per-position importance-ratio decomposition. Taking the minimum over all bounds yields the tightest known guarantee across all divergence regimes. Crucially, all bounds depend on the maximum token-level divergence $D_{mathrm{KL}}^{mathrm{tok,max}}$ (or $D_{mathrm{TV}}^{mathrm{tok,max}}$), a sequence-level quantity that cannot be controlled by token-independent methods like PPO clipping. We propose Trust Region Masking (TRM), which masks entire sequences violating the trust region, enabling the first non-vacuous monotonic improvement guarantees for long-horizon LLM-RL.