#01Jul 27, 2026
cs.AI
Eviction as Estimation: A Fixed-Lag Smoothing View of Test-Time Memory, and When Measuring Beats Accumulating
Maruthi Vemula, Neeraj Praneeth Gajula
A language model with a bounded working memory must repeatedly decide which stored items to keep. Every deployed method decides the moment an item arrives, from the past (StreamingLLM, H2O) or from a guess about the future (SnapKV). We recast the choice as an estimation problem on a hidden signal, whether an item will be reused, placing existing methods on one axis, the commit lag $H$: online filters and learned predictors commit at $H=0$, while Belady's offline optimum sits where the whole future is known. The missing regime in between, fixed-lag smoothing, waits a bounded number of steps, observes which items a correct near-future prediction attended to, and only then commits. This measurement, demonstrated utility, turns Belady's unobservable future request into something we read off the model itself. We instantiate it as a training-free policy, RMM, a strict generalization of H2O that reduces to it exactly when the measurement is uniform. In controlled settings where reuse is endogenous and separated in time, demonstrated utility identifies used memory far better than accumulated attention, and a small bounded memory behaves like a much larger one. But on independent third-party benchmarks, run inside NVIDIA's KVPress harness against its own SnapKV, H2O, and StreamingLLM implementations, the advantage mostly disappears: RMM is on par with H2O for single-turn question answering and loses to both H2O and SnapKV in a streaming multi-turn setting. The cause is simple: on natural text the model is correct about most tokens, so weighting attention by correctness barely changes it, and demonstrated utility collapses onto accumulated attention unless reuse is sharp and endogenous, which standard benchmarks do not exercise. Our contribution is the framework and an honest map of when measuring beats accumulating, not a new state of the art.
#02Jul 27, 2026
cs.CV
SADe: Sparse-Atom Support Decontamination for Few-Shot Segmentation with Weak Support Annotations
Hang Xing, Guangjun Liu, Yan Xia and 1 more
Few-shot segmentation (FSS) commonly assumes clean pixel-level support masks, yet practical support supervision often uses boxes, scribbles, coarse masks, or pseudo-masks. These weak annotations may include texture-similar distractors and background context alongside the target, contaminating class prototypes or visual prompts before query prediction. We introduce SADe, a predictor-agnostic support decontamination layer that estimates the reliability of selected support patches without query information. Central to SADe is sparse autoencoder (SAE) atom evidence: dense similarity may respond to both target and texture-similar context, whereas contrasting atom activations inside and outside the weak-support region provides factor-level reliability cues. A lightweight router combines atom evidence with dense similarity and episode statistics to predict patch reliability and generate a cleaned support mask. Trained once on synthetic weak-support episodes from FSS-1000, the router is frozen for all target evaluations. The resulting mask supports standalone prediction or can be supplied to heterogeneous FSS models through native support interfaces without altering query-side inference. Under a matched weak-support protocol, SADe achieves the highest query mIoU in six of nine standalone prompt-shot combinations. With the same ProMi query head, it is within 0.03 mIoU of SAM3-derived masks under tight boxes and surpasses them by 11.17 and 19.49 points under box-r2 and box-r4, respectively. As a plug-in, SADe improves over raw support in 70 of 72 matched box-family comparisons across four frozen downstream models and two datasets. On point and scribble prompts, its average performance remains close to the corresponding raw-support baseline. Ablations and atom-removal controls show that atom evidence contributes reliability information beyond dense similarity.
#03Jul 27, 2026
cs.CL
The Physics of Multi-Turn Long-Horizon Planning: From Pre-training to Post-training via Single- and Multi-Teacher On-Policy Agentic Distillation
Tianyi Men, Zhuoran Jin, Kang Liu and 1 more
Multi-turn long-horizon planning is critical for foundation model agents, yet how to fundamentally improve it remains unclear. Existing models are trained on uncontrollable and opaque Internet data, making it difficult to identify how planning ability is acquired, shaped, and integrated. To address this challenge, we introduce a unified and controlled multi-turn environment that enables precise control. It allows systematically study long-horizon planning across three stages. (1) Planning ability acquisition during pre-training. We study data format, distribution, and quality. Explicit world model construction through CoT state transition modeling yields stronger long-horizon generalization. Atomic skills alone are insufficient for compositional generalization, whereas a litte long-horizon data works. Moreover, suboptimal trajectories severely impair performance because errors amplify over long horizons. (2) Planning ability shaping via GRPO and OPD post-training. Through mutual information, we distinguish general planning patterns from task-specific planning knowledge. For planning patterns, we identify three application regions of post-training: unnecessary, effective, and unsupported. OPD has a broader effective region than GRPO under low-quality and long-horizon settings, as it provides more consistent update directions. For planning knowledge, distilling unseen procedures from a teacher with different knowledge may impair student's prior world modeling without fully establishing new knowledge. (3) Planning ability integration through MOPD post-training. We show that multi-teacher on-policy distillation (MOPD) integrates capabilities by converging to shared planning-pattern across environments. Compatible patterns enable cross-environment generalization, partially shared patterns support continual learning, while completely conflicting patterns cause severe interference.
#04Jul 27, 2026
cs.NI
Denial of Deadline: Network-Driven Accuracy Collapse in Distributed Inference Pipelines
Jhonatan Tavori, Gur-Eyal Sela, Ion Stoica and 1 more
Inference systems increasingly combine a fast path that returns predictions within the application's latency deadline together with a higher-accuracy slow path that runs higher-compute methods on stronger, remote hardware, so its results can be returned on time and combined with the fast path predictions. Across several application domains, we abstract this inference architecture as a fast path, a slow path, and a coordination layer with two functions: a router that invokes the slow path and a merger that decides whether to incorporate its returned predictions. In this work, we show that this new coordination layer exposes a new attack surface: shaped workload attacks, e.g., Yo-Yo bursts, can exploit contention at shared resources along the slow path to push benign users' slow-path predictions past their latency deadlines. The merger then discards those predictions, while the fast path continues to return timely outputs. We refer to the resulting loss of slow-path accuracy benefits as accuracy collapse. We demonstrate accuracy collapse in a two-tier edge-cloud multi-object tracking pipeline in autonomous driving. In simulation, approximately 4,000 burst-shaped requests increase benign p99 latency from 92ms to 2s, nearly eliminating the benefit of the slow path's cloud inference, reducing object tracking quality by 7.0 HOTA points on average. We further find that accuracy degradation can significantly vary (2.0-18.7 HOTA points), depending on the video intervals that are targeted in the attack, and that certain rare classes (e.g., stop signs) lose nearly half of their pre-attack prediction accuracy. These results show that workload attacks can degrade prediction quality without needing either access to model weights or victim data, and motivate research on attacks and defenses for routing, merging, scheduling, and resource isolation in these emerging inference pipeline architectures.
#05Jul 27, 2026
physics.acc-ph
A corrective agentic hybrid RAG and an operations-grounded evaluation for a scientific facility
Rajat Sainju, Dariusz Jarosz, Hairong Shang and 5 more
Scientific user facilities accumulate decades of operational knowledge that no single search index covers: electronic logbooks, technical documents, internal wikis, operations chat messages, maintenance records, and live control-system data. We present APS-RAG, Advanced Photon Source Retrieval Augmented Generation, a deployed platform that makes the institutional knowledge at the Advanced Photon Source (APS) accessible to staff through natural-language queries, along with an operations-grounded evaluation. The retrieval engine fuses dense, sparse, and knowledge-graph (KG) channels with query-type-adaptive reciprocal-rank fusion, adds a corrective agentic loop, and runs a native-tool ReAct executor over a Model Context Protocol (MCP) tooling layer. We construct APS-Bench, a 50-question, question-answering (QA) dataset with auditable gold answers. Every retrieval-augmented variant numerically improves strict vital-nugget recall over a naive BM25 baseline (63.8%), with the full corrective Agentic GraphRAG scoring (70.3%). The cross-encoder reranker contributes significantly to answer quality: removing it and allowing the LLM to score relevance drastically reduces strict vital recall by 32.8%. The graph channel and corrective loop contribute positively as expected, but the performance gains are marginal. Additionally, we also compare the performance of open-source and closed-source LLMs in final answer synthesis. We release the APS-Bench construction methodology, the six-layer evaluation harness, and the underlying codebase, along with the '/aps-rag' retrieval agent skill framework, to support reproduction and adoption at other facilities. Together, the deployed platform and its operations-grounded evaluation present a promising workflow for trustworthy, statistically grounded AI assistance in facility operations, transferable to other large scientific instruments.