Long-term memory enables large language models (LLMs) to preserve and reuse information across interactions, but it can also turn localized errors into persistent risks. Existing work mainly evaluates whether memory systems store and retrieve information correctly, leaving limited understanding of how errors propagate across responses, memory states, and future interactions. We propose a structural causal model (SCM)-based framework for cross-turn error propagation in memory-augmented LLMs. We model user questions, model responses, and memory states as a dynamic causal process, and identify two entry pathways: internal memory updating and external question feedback. By intervening on these pathways, we construct four counterfactual trajectories and quantify their downstream effects and interaction. Error influence is evaluated at four levels: memory retention, natural responses, targeted diagnostic probing, and probability-level error preference. Experiments show that error influence generally decays with interaction distance, while the memory-update pathway contributes more persistent effects than question feedback; latent errors may remain even after disappearing from natural responses. Propagation patterns also vary across memory categories and memory mechanisms. Pathway-guided restoration further validates this decomposition: Question Repair reduces residual error by 27.5%, Memory Repair by 70.2%, and Joint Repair by 98.3%, nearly eliminating residual propagation.
Tool-using agents fail two ways: choosing the wrong tool, or forming wrong arguments, and an early failure of either kind can silently corrupt everything downstream. We measure a correct-invocation rate that separates the two, under both a clean teacher-forced context and the model's own free-running context, on five open-weight models over contamination-free multi-step tasks (depths 1-8). By depth 6, roughly 70% of a model's own clean-context capability is lost to its own earlier mistakes (L6 = 0.686, 0.684). Our central finding concerns the measurement itself. Under exact-match scoring against a fixed gold trajectory, a propagation model's severity and recovery parameters are not merely hard to estimate - they are fixed by the scoring rule. Severity is forced to its boundary (0 of 869 poisoned steps correct); recovery is structurally unobservable (0 of 580 poisoned steps returned on-track, against an expected 0.0058 by chance). Both follow from one mechanism: post-divergence, the gold value is generated by tool constants the model never sees, so it is information the model cannot derive. A fit run anyway returns 0.92 and 0.73 for a quantity that is exactly 1.000 - confident numbers for a parameter the scoring rule already determined. We give the mechanism and a remedy, conditional-on-state scoring, applied retrospectively to cached completions at zero additional cost, which un-pins severity to interior estimates excluding zero (+0.149, +0.316).
Yaokun Liu, Yifan Liu, Daniel Yue Zhang +3cs.MA cs.CL
LLM-based multi-agent systems (MAS) solve complex tasks through communication among role-specialized agents. However, inter-agent dependencies introduce reliability risks beyond isolated agent failures. For instance, errors in intermediate messages could be inherited and amplified by downstream agents. Existing uncertainty quantification (UQ) methods mainly target isolated responses or single-agent reasoning, and therefore fail to capture uncertainty propagation in MAS. To this end, we propose PropUQ-MAS, an error propagation-aware UQ framework that represents MAS execution as a communication-structured graph and estimates each step's reliability by combining local uncertainty with uncertainty inherited from upstream messages. Extensive experiments demonstrate that PropUQ-MAS consistently improves UQ in MAS, with average relative gains of +6.10% in AUROC and +47.58% in PRR.
Mohanad Odema, Gabrielle De Micheli, Dayin Gou +3cs.AI cs.CL cs.PF
Training-free low-rank compression frameworks have been gaining prominence for LLM compression given their effectiveness in reducing model parameter count while maintaining task-level accuracy. However, existing SOTA frameworks share two key limitations: (1) residual errors in calibration data activations accumulate across layers during compression, causing misalignment between representations simulated at compression time and those experienced at inference; (2) the assumption that layer importance distribution is preserved post-compression does not hold. Together, these two effects introduce misalignment in the compression process in relation to the deployed model. We study these effects and propose a simple, training-free methodology compatible with existing frameworks to mitigate them, comprising: (1) Layer-by-Layer Compression with Calibration Correction; (2) Iterative Compression with Rank Allocation Correction. Implemented atop an existing SOTA decomposition framework, and evaluated on Llama and Qwen3 models across various benchmarks and compression rates, our approach demonstrates up to ~1-2.5 accuracy point improvements over per-weight and joint decomposition baselines on zero-shot tasks.
Chenxi Zhou, Pengfei Cao, Jinyu Ye +5cs.AI cs.CL cs.LG
Post-training quantization reduces the deployment cost of large language models, yet how severely a quantized model degrades is not determined by bit-width alone. We systematically study weight-only post-training quantization across bit-widths, quantization methods, model scales and downstream tasks on multiple model families. We observe that such degradation varies substantially across these factors: 4-bit quantization usually preserves performance, 2-bit often causes broad degradation, and at 3-bit, degradation becomes apparent but varies markedly with task type, quantization method and model scale. To explain this variability, we use the signal-to-noise ratio (SNR) to measure how strongly quantization perturbs full-precision representations. We trace degradation back to two linked processes: how quantization errors arise within individual modules, and how they accumulate across layers. First, a source SNR decomposition shows that newly introduced errors depend on three factors: the magnitude of the weight error, the strength of the task-specific signal, and how strongly the quantization error aligns with task-specific activations. Different factors affect these components in distinct ways. Second, a cross-layer propagation analysis shows that these errors can be attenuated, preserved, or amplified as they pass across layers, and that larger models benefit from weaker error amplification. Together, these results establish that quantization degradation is governed by how errors are introduced at the source and how they accumulate across the network.
This paper explores the "Granularity Paradox" in time-series forecasting, wherein finer temporal disaggregation (e.g., Monthly to Weekly/Daily) improves in-sample diagnostics and dataset size (N), but degrades out-of-sample accuracy due to recursive error compounding over longer horizons (H). Conversely, coarse aggregation (Annual) eliminates recursive error propagation but reduces data available to estimators. We formalize this trade-off and benchmark 10 models - spanning naïve, statistical, machine learning, and deep learning architectures - across six granularities using a 13-year public procurement dataset. The empirical results reveal a non-monotonic threshold structure: recursive autoregressive and seasonal models degrade substantially under high-frequency forecasting (e.g., Holt-Winters reaches a Test R-squared of -151 and TPFE of 425.85% at the Daily grain), while the LSTM traces a U-shaped error curve, worsening from Monthly (19.66%) through Bi-Weekly (35.94%) before overcoming the error propagation penalty at Daily (TPFE of 4.35%, R-squared of 0.66). Linear Regression remains stable across all granularities (16.3-17.0% TPFE), confirming that the paradox is driven by recursive feedback topology, not model complexity. The results demonstrate that standard pointwise metrics (RMSE, MAE) systematically mask cumulative error propagation, and that evaluating forecasts without goal-dependent cumulative metrics produces misleading assessments of model adequacy. We introduce a consensus-dissensus diagnostic comparing the directional behaviour of pointwise metrics against cumulative TPFE across granularities, enabling the identification of models whose standard diagnostics mask systematic error propagation.
Multi-agent AI systems can improve answer selection by allowing different language models to exchange reasoning traces, revise initial predictions, and support a final decision. However, such communication may also introduce reliability risks: reasoning from one agent can correct another agent's mistake, but it can also mislead an agent that was initially correct. This paper studies reliable multi-agent AI communication through reasoning exchange and runtime answer revision. We develop a framework in which agents first answer multiple-choice questions independently, then share reasoning traces and revise their decisions. We conduct numerical experiments where we evaluate whether this process improves accuracy, produces more positive than negative answer transitions, and remains effective across domains such as cybersecurity, networking, and general knowledge. The results help identify when multi-agent reasoning improves reliability and when it may propagate errors.
Sequential multi-agent LLM pipelines chain specialized agents without verification at handoffs, creating a structural flaw with measurable and severe consequences. We show that hallucinations injected at Stage 1 do not merely persist; they transform: raw numerical facts become derived computations, then narrative prose, then editorially approved conclusions. At each transformation, detectability degrades near-irreversibly. We formalize this as the hallucination snowball effect, a first-order Markov process over four states (Raw Fact $\to$ Derived $\to$ Narrative $\to$ Invisible) with empirically measured per-boundary escape probabilities of 24.6%, 48.3%, and 89.3%. Across 346 automatically injected hallucinations in a 4-agent financial analysis pipeline on FinanceBench, gpt-4o detection drops from 72.0% at Stage 1 to 50.9% at Stage 4, and 23.7% of hallucinations survive completely undetected in the final output. Even the strongest model tested (Qwen3.5-397B-A17B, 87.0% at Stage 1) faces a structural ceiling; projected Stage 4 detection is only ${\sim}$60--65%. Critically, boundary gates using identical RAG verification tools reduce hallucination survival from 58.4% to 16.2% versus end-of-pipeline checking (Cohen's $h = -0.911$, $p < 0.000001$), while end-checking alone achieves merely 2.3 pp improvement over no verification. When you verify matters more than whether you verify. Our model predicts survival for $n$-agent linear pipelines and prescribes optimal verification resource allocation: invest at $S_1{\to}S_2$ first, where 75.4% of hallucinations are still catchable, not at $S_3{\to}S_4$ where 89.3% have already escaped.
Modular neural network pipelines suffer from error compounding: noise at any module boundary propagates and potentially amplifies through subsequent modules. We introduce energy conservation as a hard physical constraint on inter-module information flow. Activation energy (the squared L2 norm of feature vectors) is enforced to be exactly preserved at every module boundary. Unlike soft energy penalties, conservation is an inviolable law: the network may redistribute energy across neurons but cannot create or destroy it. Four experiments on CIFAR-10 demonstrate: (1) conservation retains 77.4% of clean accuracy at noise sigma=0.2, versus 35.1% for baselines and 30.9% for energy-penalized models (p<0.001, 5 seeds); (2) pipelines become depth-invariant, retaining 93.3% at depths 2 through 5 with noise at every boundary; (3) the advantage generalizes to systematic bias (+45.1%), Gaussian (+40.4%), and adversarial noise (+4.8%), with a principled non-effect on dropout (-0.3%); (4) on ResNet-18, the conservation advantage scales inversely with intrinsic normalization: +0.3 pp with BatchNorm, +26.2 pp without at sigma=0.2, reaching +58.0 pp at sigma=0.5. Experiment 5 validates the operator on a real modular robotic pipeline (MuJoCo physics, Franka Panda). Across three independent runs on separate machines (90 trials per cell), conservation provides +18.9 pp average advantage on monocular-depth-style noise. A formal bound proves conserved noise energy is strictly less than input noise energy.
Multi-step agentic retrieval-augmented generation (RAG) pipelines have demonstrated significant capability for complex reasoning tasks, yet remain vulnerable to a class of failure that existing hallucination detection mechanisms systematically miss: cascading hallucination, where errors introduced at early pipeline stages propagate and amplify across successive reasoning steps, producing confident but factually incorrect final outputs. To address this vulnerability, we formalize cascading hallucination as a distinct failure mode in agentic RAG systems, present a four-type taxonomy of cascade patterns, and introduce CHARM (Cascading Hallucination Aware Resolution and Mitigation), an architectural framework for detecting and interrupting error propagation in multi-step reasoning pipelines. CHARM comprises four components - stage-level fact verification, cross-stage consistency tracking, confidence propagation monitoring, and cascade resolution triggering - that operate alongside standard agentic RAG pipelines without requiring architectural replacement. We evaluate CHARM on HotpotQA, MuSiQue, 2WikiMultiHopQA, and a custom adversarial dataset across LangChain agentic pipeline configurations, achieving an 89.4% cascade detection rate with a 5.3% false positive rate and 215 ms +/- 18 ms average latency overhead per stage, achieving an error propagation reduction of 82.1%, compared to 18.5% for output-level detectors. Component ablations confirm that each detection module contributes meaningfully to overall cascade coverage. CHARM integrates with human-in-the-loop oversight frameworks to provide a complete reliability and governance stack for production agentic AI deployment.