# When Adaptive Rewards Hurt: Causal Probing and the Switching-Stability Dilemma in LLM-Guided LEO Satellite Scheduling

Source: [When Adaptive Rewards Hurt: Causal Probing and the Switching-Stability Dilemma in LLM-Guided LEO Satellite Scheduling](<https://www.sota2.com/research/paper/69d48fa77d0172140ec74441>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-08T05:00:04.658Z

Authors: Yuanhang Li

Published: 2026-04-04T03:04:53.000Z

## Abstract

Adaptive reward design for deep reinforcement learning (DRL) in multi-beam LEO satellite scheduling is motivated by the intuition that regime-aware reward weights should outperform static ones. We systematically test this intuition and uncover a switching-stability dilemma: near-constant reward weights (342.1 Mbps) outperform carefully-tuned dynamic weights (103.3+/-96.8 Mbps) because PPO requires a quasistationary reward signal for value function convergence. Weight adaptation-regardless of quality-degrades performance by repeatedly restarting convergence. To understand why specific weights matter, we introduce a single-variable causal probing method that independently perturbs each reward term by +/-20% and measures PPO response after 50k steps. Probing reveals counterintuitive leverage: a +20% increase in the switching penalty yields +157 Mbps for polar handover and +130 Mbps for hot-cold regimes-findings inaccessible to human experts or trained MLPs without systematic probing. We evaluate four MDP architect variants (fixed, rule-based, learned MLP, finetuned LLM) across known and novel traffic regimes. The MLP achieves 357.9 Mbps on known regimes and 325.2 Mbps on novel regimes, while the fine-tuned LLM collapses to 45.3+/-43.0 Mbps due to weight oscillation rather than lack of domain knowledge-output consistency, not knowledge, is the binding constraint. Our findings provide an empirically-grounded roadmap for LLM-DRL integration in communication systems, identifying where LLMs add irreplaceable value (natural language intent understanding) versus where simpler methods suffice.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Traffic Regime Adaptation on Novel Regimes (IoT burst, polar handover, hot-cold) (test)](<https://www.sota2.com/research/sota/traffic-regime-adaptation-on-novel-regimes-iot-burst-polar-handover-hot-cold-test.md>): Traffic Regime Adaptation; Novel Regimes (IoT burst, polar handover, hot-cold) (test); Novel Rate (Mbps): 342.1.

- [Satellite Traffic Management on Known Traffic Regimes](<https://www.sota2.com/research/sota/satellite-traffic-management-on-known-traffic-regimes.md>): Satellite Traffic Management; Known Traffic Regimes; Throughput (Mbps): 357.9.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.03562>)

[PDF](<https://arxiv.org/pdf/2604.03562.pdf>)
