Drift No More: Spectral Locking for Long-Horizon Open-Weight Multi-Agent AI
Organizations: University of Florida · University of Oklahoma · SRI International
Abstract
Long-horizon multi-agent systems suffer from three tightly coupled failures: task drift, role drift, and state staleness. We test the hypothesis that these share a common representational cause: agent hidden states move away from the directions encoding the active task, assigned role, and current shared state. A pilot study over 14,832 annotated turns supports this: drifted turns collapse near-uniformly across all three anchor subspaces (pairwise r = 0.72-0.78). Existing methods address these failures separately, operating at the surface-text level or treating control directions independently. We introduce ATLAS (Algebraic Task-role Locking with Adaptive Spectral Memory), a training-free intervention acting on hidden states before a token is generated. ATLAS builds a positive-semidefinite anchor kernel over task, role, and state subspaces and steers hidden states toward their shared high-coherence eigendirections; a low-rank Gram decomposition keeps this tractable. Across long-horizon memory, state maintenance, tool use, least-privilege coordination, and embodied stress tests, ATLAS outperforms prompting, retrieval, memory, and activation-steering baselines. Against A-MEM it improves memory macro-average accuracy by up to 5.7 points and reduces write/read failures by up to 31.7%/24.8%; against MemoryArena-Agent it gains up to 4.8 points. It gains 10.7-12.1 points under perturbation, reaches 93.5% role consistency, and transfers to a 72B backbone (macro-average 61.0 vs. 56.2). Ablating the joint task-role-state span costs 7.2 points against 0.8 for a matched-rank random span, indicating these directions are functionally implicated, not incidental; supplying them by prompt or retrieval recovers 4.3 points less. ATLAS is therefore a representation-level controller for open-weight multi-agent systems, complementary to output-level protocols for black-box agents.
Figures & tables
| T1: Homogeneous-10 ( ) | T2: Homogeneous-12 ( ) | T3: Heterogeneous-12/6M ( ) | ||||||||||
| FGR | MA | PU | Avg | FGR | MA | PU | Avg | FGR | MA | PU | Avg | |
| Single-Agent | 41.2 | 33.8 | 29.4 | 34.8 | 43.5 | 35.1 | 31.2 | 36.6 | 38.7 | 31.6 | 27.8 | 32.7 |
| Prompt-Only | 46.3 | 38.2 | 34.5 | 39.7 | 48.1 | 39.7 | 36.0 | 41.3 | 43.8 | 35.9 | 32.1 | 37.3 |
| Struct-Role | 49.1 | 40.6 | 36.8 | 42.2 | 51.4 | 42.3 | 38.4 | 44.0 | 46.5 | 38.4 | 34.6 | 39.8 |
| RAG-Mem | 52.7 | 43.1 | 38.9 | 44.9 | 54.6 | 44.8 | 40.5 | 46.6 | 49.8 | 41.0 | 36.7 | 42.5 |
| A-MEM | 54.3 | 45.7 | 40.2 | 46.7 | 56.1 | 47.2 | 41.8 | 48.4 | 51.5 | 43.4 | 38.1 | 44.3 |
| T1: Homogeneous-10 ( ) | T2: Homogeneous-12 ( ) | T3: Heterogeneous-12/6M ( ) | |||||||||||||
| OA | NMS | W | R | U | OA | NMS | W | R | U | OA | NMS | W | R | U | |
| Single-Agent | 38.1 | 34.6 | 28.3 | 21.5 | 12.1 | 40.7 | 37.2 | 26.1 | 20.3 | 12.9 | 35.4 | 32.1 | 30.5 | 22.8 | 11.3 |
| Prompt-Only | 43.5 | 40.2 | 24.7 | 19.2 | 12.6 | 45.9 | 42.5 | 22.8 | 18.1 | 13.2 | 40.6 | 37.8 | 26.9 | 20.4 | 12.1 |
| Struct-Role | 46.2 | 42.8 | 22.4 | 18.5 | 12.9 | 48.3 | 45.1 | 20.6 | 17.3 | 13.8 | 43.5 | 40.1 | 24.5 | 19.6 | 12.4 |
| RAG-Mem | 48.9 | 45.4 | 21.1 | 16.8 | 13.2 | 51.2 | 47.9 | 19.3 | 15.7 | 13.8 | 46.1 | 42.5 | 23.1 | 18.2 | 12.6 |
| A-MEM | 51.3 | 48.1 | 18.5 | 15.2 | 15.0 | 53.8 | 50.6 | 16.7 | 14.1 | 15.4 | 48.7 | 45.3 | 20.4 | 16.8 | 14.1 |
| Variant | Avg | |
|---|---|---|
| Full ATLAS | 54.1 | — |
| Component (leave-one-out) | ||
| w/o role stabilization | 49.1 | 5.0 |
| w/o state refresh | 48.9 | 5.2 |
| w/o task-switch correction | 51.2 | 2.9 |
| w/o steering | 52.1 | 2.0 |
Appendix figures & tables21 assets
Supplementary material from the paper’s appendix.
Appendix
| Agent | Backbone | Purpose |
|---|---|---|
| Coordinator | Llama-4-Scout-17B | routing, objective tracking |
| Planner | Qwen3-8B | subgoal decomposition |
| Memory-Writer | Qwen2.5-7B-Inst. | state insertion & update |
| Memory-Reader | Qwen2.5-7B-Inst. | state retrieval |
| Executor | Phi-4-14B | final response generation |
| Verifier | Gemma-2-9B-Inst. | consistency checking |
| Variant | FGR | MA | PU | Avg | |
|---|---|---|---|---|---|
| Full ATLAS | 61.7 | 53.1 | 47.6 | 54.1 | — |
| w/o role stabilization | 56.8 | 48.3 | 42.1 | 49.1 | |
| w/o state refresh | 57.2 | 45.6 | 43.8 | 48.9 | |
| w/o task-switch correction | 58.4 | 50.2 | 44.9 | 51.2 | |
| w/o steering | 59.1 | 51.4 | 45.7 | 52.1 | |
| w/o Monitor agent | 57.9 | 47.8 | 43.2 | 49.6 |
| Method | AST | AC | MC |
|---|---|---|---|
| Single-Agent | 74.6 | 61.7 | 42.1 |
| Prompt-Only | 78.3 | 65.4 | 47.8 |
| Struct-Role | 80.1 | 67.9 | 50.5 |
| RAG-Mem | 82.5 | 69.8 | 53.2 |
| A-MEM | 83.9 | 70.2 | 54.3 |
| MemoryArena-Agent | – | – | – |
| Method | TS | PV | OE |
|---|---|---|---|
| Single-Agent | 54.2 | 18.7 | 23.4 |
| Prompt-Only | 58.6 | 15.3 | 19.8 |
| Struct-Role | 63.8 | 12.1 | 16.5 |
| RAG-Mem | 62.4 | 13.5 | 17.2 |
| A-MEM | 65.1 | 11.2 | 14.6 |
| MemoryArena-Agent | – | – | – |
| M1: Llama+Qwen3VL | M2: Qwen3+Qwen2.5VL | M3: Phi4+Phi4V | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Method | SR | PE | SP | SR | PE | SP | SR | PE | SP |
| Single-Agent | 72.3 | 64.8 | 18.5 | 75.1 | 67.3 | 20.2 | 68.9 | 61.4 | 15.7 |
| Prompt-Only | 78.6 | 70.2 | 24.3 | 81.2 | 73.1 | 27.1 | 75.1 | 66.9 | 21.4 |
| Struct-Role | 81.4 | 73.6 | 29.7 | 84.0 | 76.2 | 32.5 | 78.3 | 70.1 | 26.2 |
| RAG-Mem | 83.7 | 75.9 | 33.1 | 86.4 | 78.5 | 36.4 | 80.8 | 72.4 | 29.8 |
| A-MEM | 85.2 | 77.4 | 36.8 | 87.9 | 80.1 | 39.6 | 82.1 | 74.3 | 33.4 |
| Method | Role-Conf. | State-Del. | Task-Sw. |
|---|---|---|---|
| Prompt-Only | 34.1 | 12.8 | 34.4 |
| Struct-Role | 41.3 | 16.5 | 39.7 |
| RAG-Mem | 44.8 | 20.6 | 43.8 |
| A-MEM | 48.2 | 24.1 | 46.5 |
| ActAdd | 42.5 | 19.3 | 42.6 |
| SEA | 44.6 | 22.3 | 40.8 |
| Variant | FGR | MA | PU | Avg |
| Prompt ATLAS | 56.4 | 47.5 | 42.8 | 48.9 |
| Retrieval ATLAS | 57.2 | 48.6 | 43.6 | 49.8 |
| Random-subspace ATLAS | 55.0 | 46.8 | 41.7 | 47.8 |
| Orthogonalized ATLAS | 58.4 | 48.9 | 43.6 | 50.3 |
| Hidden ATLAS | 61.7 | 53.1 | 47.6 | 54.1 |
| vs. best non-hidden control |
| T1: Homogeneous-10 | T2: Homogeneous-12 | T3: Heterogeneous-12/6M | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Method | RC | SF | TSR | TDA | RC | SF | TSR | TDA | RC | SF | TSR | TDA |
| Prompt-Only | 71.3 | 52.4 | 0.61 | 0.38 | 73.8 | 54.9 | 0.64 | 0.35 | 68.5 | 49.7 | 0.57 | 0.41 |
| Struct-Role | 79.6 | 56.1 | 0.67 | 0.31 | 82.1 | 58.7 | 0.70 | 0.28 | 76.9 | 53.4 | 0.63 | 0.34 |
| RAG-Mem | 76.4 | 62.8 | 0.69 | 0.29 | 78.9 | 65.3 | 0.72 | 0.26 | 73.8 | 59.7 | 0.65 | 0.32 |
| A-MEM | 78.2 | 67.5 | 0.72 | 0.26 | 80.7 | 70.1 | 0.75 | 0.23 | 75.6 | 64.8 | 0.68 | 0.29 |
| MemArena-Ag. | 79.1 | 69.0 | 0.73 | 0.24 | 82.0 | 72.4 | 0.77 | 0.21 | 77.2 | 66.1 | 0.70 | 0.27 |
| Method | Tok. | Calls | Lat. | C-N |
|---|---|---|---|---|
| Single-Agent | 28.4 | 1.0 | 4.2 | 1.29 |
| Prompt-Only | 84.7 | 10.3 | 14.6 | 0.49 |
| RAG-Mem | 107.3 | 14.1 | 19.3 | 0.43 |
| A-MEM | 118.5 | 16.7 | 22.1 | 0.41 |
| MemArena-Ag. | 112.6 | 15.4 | 20.8 | 0.44 |
| SEA | 91.2 | 10.8 | 16.8 | 0.51 |
| FGR | MA | PU | Avg | Tok. | C-N | |
|---|---|---|---|---|---|---|
| 1 | 43.5 | 35.1 | 31.2 | 36.6 | 28.4 | 1.29 |
| 3 | 54.3 | 45.6 | 40.1 | 46.7 | 58.2 | 0.80 |
| 5 | 58.1 | 49.8 | 44.3 | 50.7 | 74.5 | 0.68 |
| 10 | 60.8 | 52.4 | 46.5 | 53.2 | 90.1 | 0.59 |
| 12 | 61.7 | 53.1 | 47.6 | 54.1 | 96.8 | 0.56 |
| Measure | Coherent | Drifted |
|---|---|---|
| (task) | 0.82 | 0.45 |
| (role) | 0.80 | 0.41 |
| (state) | 0.79 | 0.44 |
| (all) | 0.81 | 0.43 |
| Pairwise drift correlation (Pearson ) | ||
| task role | 0.78 | |
| Kernel configuration | Avg. |
|---|---|
| Task only ( ) | 46.8 |
| Role only ( ) | 47.6 |
| State only ( ) | 48.3 |
| 50.4 | |
| 49.1 | |
| 51.6 |
| Method | FGR | MA | PU | Avg. |
|---|---|---|---|---|
| A-MEM | 63.4 | 54.1 | 48.5 | 55.3 |
| MemoryArena-Agent | 64.1 | 55.0 | 49.5 | 56.2 |
| ATLAS | 69.0 | 59.8 | 54.2 | 61.0 |
| Comparison | Turns | Agreement | Cohen’s |
|---|---|---|---|
| Human 1 vs. Human 2 | 200 | 91.0% | 0.82 |
| Human consensus vs. model majority | 200 | 89.5% | 0.79 |
| Human consensus vs. GPT-4o RC judge | 200 | 88.5% | 0.76 |
| Hidden-state condition | Avg. | Change |
|---|---|---|
| No removal | 54.1 | – |
| Matched-rank random removal | 53.3 | |
| Remove task–role–state span | 46.9 |
| Anchor condition | Avg. | |
|---|---|---|
| Original anchors | 54.1 | – |
| Paraphrased, 3 sets | 53.8 | |
| 75% length | 53.6 | |
| 50% length | 52.4 | |
| 25% length | 49.2 | |
| 30% lexical noise | 50.1 |
| Parameter | Alternative settings | EverMemBench Avg. |
|---|---|---|
| 0.05 / 0.10 / 0.20 | 51.4 / 53.2 / 53.6 | |
| 0.90 / 0.98 | 52.7 / 54.3 | |
| 0.05 / 0.15 | 53.0 / 53.8 | |
| 0.5 / 1.5 | 52.9 / 54.0 | |
| 0.85 / 0.95 | 53.4 / 54.2 | |
| 0.10 / 0.20 | 53.5 / 53.9 |
| Method | Midpoint | Per-backbone | Gain |
|---|---|---|---|
| ActAdd | 40.4 | 41.0 | |
| SEA | 41.3 | 42.0 | |
| ATLAS | 47.9 | 49.2 |
Multi-Agent LLM Systems
Multi-agent large language model systems coordinate multiple AI agents to solve tasks through communication, specialization, shared tools, and memory.
See all 1,100 papers