More Is Not Always Better: Cross-Component Interference in LLM Agent Scaffolding
For practitioners building LLM agent systems, this work demonstrates that the common assumption of 'more components is better' is flawed, and provides evidence for task-specific subset selection over maximally-equipped defaults.
LLM agent systems suffer from cross-component interference (CCI), where adding more scaffolding components (planning, tools, memory, etc.) degrades performance. On HotpotQA, a single-tool agent outperforms the All-In system by 32% (F1 0.233 vs 0.177), and on GSM8K, a 3-component subset beats All-In by 79% (0.43 vs 0.24).
LLM agent systems are built by stacking scaffolding components (planning, tools, memory, self-reflection, retrieval) assuming more is better. We study cross-component interference (CCI): degradation when components interact destructively. We run a full factorial experiment over all 2^5=32 subsets of five components on HotpotQA and GSM8K with Llama-3.1-8B/70B (96 conditions, up to 10 seeds). The All-In system is consistently suboptimal: on HotpotQA, a single-tool agent surpasses All-In by 32% (F1 0.233 vs 0.177, p=0.023); on GSM8K, a 3-component subset beats All-In by 79% (0.43 vs 0.24, p=0.010). Optimal component count is task-dependent (k*=1-4) and scale-sensitive: at 70B, combinations that hurt at 8B provide gains, though All-In still trails the best subset. We fit a main-effects regression (R^2=0.916, adj-R^2=0.899, LOOCV=0.872), compute exact Shapley values, and find 183/325 submodularity violations (56.3%), showing greedy selection is unreliable. A three-body synergy among Tool Use, Self-Reflection, and Retrieval (INT_3=+0.175, 95% CI [+0.003,+0.351]) is reported as exploratory. CCI replicates across model families (Qwen2.5) and is robust to prompt paraphrasing. Our findings suggest maximally-equipped agent defaults should be replaced by task-specific subset selection via interaction-aware analysis.