NeurIPS 2026 Papers

Vision Language Models (VLMs) offer the exciting possibility of processing text as rendered images, bypassing the need for tokenizing the text into long token sequences. Since VLM image encoders map fixed-size images to a fixed number of visual tokens, varying rendering resolution provides a fine-grained compression knob. However, accuracy deteriorates quickly as compression increases: characters shrink below the vision encoder's effective resolution, making them indistinguishable. To address this, we propose LensVLM, an inference framework and post-training recipe that enables VLMs to scan compressed images, then selectively expand only the relevant images to their uncompressed form via learned tools. Building on Qwen3.5-9B-Base, LensVLM maintains accuracy comparable to the full-text upper bound at 4.3$\times$ effective compression and outperforms retrieval-based, text- and visual-compression baselines up to 10.1$\times$ effective compression across seven text QA benchmarks. LensVLM also generalizes to multimodal document and code understanding tasks, with the accuracy gain over baselines growing as compression increases. Our analysis validates this approach: training makes visual compression robust to rendering choices, and as compression grows the model increasingly relies on expanded content rather than unreliable visual reading. The analysis also yields practical tool-choice guidance: text expansion is preferable for rendered text, while high-resolution image expansion suits native documents whose layout cues carry task-relevant information.

Scalar quantization of large language models (LLMs) is fundamentally limited by information-theoretic bounds. While vector quantization (VQ) overcomes these limits by encoding blocks of parameters jointly, practical implementations must avoid the need for expensive lookup mechanisms or other explicit codebook storage. Lattice approaches address this through highly structured and dense packing. This paper explores the Leech lattice, which, with its optimal sphere packing and kissing configurations at 24 dimensions, is the highest dimensional lattice known with such optimal properties. To make the Leech lattice usable for LLM quantization, we extend an existing search algorithm based on the extended Golay code construction, to i) support indexing, enabling conversion to and from bitstrings without materializing the codebook, ii) allow angular search over union of Leech lattice shells, iii) propose fully-parallelisable dequantization kernel. Together this yields a practical algorithm, namely Leech Lattice Vector Quantization (LLVQ). LLVQ delivers state-of-the-art LLM quantization performance, outperforming recent methods such as Quip\#, QTIP, and PVQ. These results highlight the importance of high-dimensional lattices for scalable, theoretically grounded model compression.

Large language models (LLMs) trained on next-token prediction exhibit remarkable in-context learning (ICL) abilities, yet the representations that support ICL remain poorly understood. We consider such representations in a controlled setting: prompting LLMs with data emitted from hidden Markov models (HMMs) and probing for the corresponding belief state -- the posterior distribution over the HMM's hidden states given the observed token history. Across six open-source LLMs prompted with data from 40 HMMs selected for non-trivial belief structure, we find that belief states are linearly decodable from residual stream activations, with peak probe $R^2$-values ranging from 0.83 to 0.99 across HMM and LLM combinations, typically occurring at early-to-middle layers. To establish functional relevance, we intervene directly on the probe-identified subspace via patching and steering, resulting in downstream prediction quality on the order of the untampered model, while controls degrade substantially. Together, these results provide representation-level evidence that ICL in pre-trained LLMs approximates optimal Bayesian prediction over a context-inferred HMM. More broadly, our findings extend prior results linking input-distribution structure to activation geometry: from toy networks trained explicitly on HMM data to production-scale transformers trained on natural language.

Transformers acquire in-context learning abilities through abrupt phases during training, often unfolding over multiple stages in which key circuits, such as induction heads, emerge. In this work, we characterize the dynamics underlying the emergence of such circuits across these stages. We focus on a synthetic associative recall task, where sequences are drawn from random maps between a permutation group and a vocabulary, and the model is required to complete the mapping of a permutation by retrieving the corresponding association from context. For this task, we study the gradient-flow trajectories of a simplified two-layer attention-only transformer. Leveraging symmetries in both the transformer architecture and the data distribution, we derive *closed-form parameter dynamics* for both attention layers. We identify a conservation law that couples parameters across layers and governs their joint learning. This law reveals how initialization controls the *timescales* over which such circuits emerge. In the vanishing-initialization limit, we characterize the gradient-flow trajectory, showing how training jumps from saddle to saddle. Finally, we provide empirical evidence across different architectural choices, validating our simplifications and extending the insights from our analysis beyond the simplified setting.

For LLM agents to be useful beyond simple chat, they must hold sensitive user context such as calendars, credentials, health records, and financial data. We study whether the mere presence of such secrets in a model's context window introduces hidden correlations into the model's benign outputs, allowing reconstruction even when the model correctly refuses direct extraction. We further study whether an adversary can actively engineer prompts that amplify this effect, using the model as a covert carrier to transmit secrets through seemingly innocuous text. In both cases, this limited leakage is exploited using a novel _adaptive_ attack that assumes black-box access to the underlying model. In controlled experiments across eight proprietary models, we find that 2-digit in-context secrets are reconstructed with near-perfect accuracy and 4-digit secrets at 82% exact match, all from outputs the model produces in response to ordinary, non-adversarial requests. More capable models leak more: stronger instruction-following amplifies sensitivity to in-context secrets, making leakage a byproduct of capability rather than a bug to be patched. We show this leakage enables two practical attacks: (1) a trained classifier that infers semantic predicates about user memories (e.g., health conditions, financial events) from routine natural-language outputs, and (2) an RL-trained adversary that extracts full Social Security Numbers from a production-style agent at 100% success.

Critical transitions – abrupt, often irreversible changes in system dynamics – arise across human and natural systems, often with catastrophic consequences. Real-world observations of such shifts remain scarce, preventing the development of reliable early warning systems. Conventional statistical and spectral indicators, such as increasing variance, tend to fail under realistic conditions of limited data and correlated noise, whereas existing deep learning classifiers do not extrapolate beyond their training data distribution. In this work, we introduce TipPFN, an in-context learning (ICL) framework that uses a prior-data fitted network to infer a system's proximity to a critical transition. Trained on our novel synthetic data generator, which is based on canonical bifurcation scenarios coupled to diverse, randomized stochastic dynamics, TipPFN flexibly capitalizes on contexts of various sizes, complexity and dimensionalities. We demonstrate robust, state-of-the-art early detection of critical transitions in previously unseen tipping regimes, sim-to-real examples, and real-world observations in both ICL and zero-shot settings.

Recently, a series of works have started studying variations of concepts from learning theory for product spaces, which can be collected under the name high-arity learning theory. In this work, we consider a high-arity variant of sample compression schemes and we prove that the existence of a high-arity sample compression scheme of non-trivial quality implies high-arity PAC learnability.

Large language model (LLM) agents have demonstrated strong capabilities in long-horizon tasks by interleaving reasoning with tool use. However, as these agents scale to complex workflows such as software engineering and open-ended research, context management becomes a fundamental bottleneck: interaction histories grow unbounded, become costly to maintain, and are difficult to reuse across sessions and agents. We introduce \textbf{Git-Context-Controller (GCC)}, a structured context management framework inspired by software version control systems. GCC elevates agent context from a transient token stream to a persistent, navigable memory workspace with explicit operations---\texttt{COMMIT}, \texttt{BRANCH}, \texttt{MERGE}, and \texttt{CONTEXT}, that enable milestone-based checkpointing, isolated exploration of alternative reasoning paths, and hierarchical retrieval of historical context. By organizing agent memory as a versioned file system, GCC allows agents to manage long-term goals, recover and transfer reasoning across sessions, and coordinate multi-trajectory problem solving in a principled manner. Empirically, agents equipped with GCC achieve state-of-the-art performance on both SWE-Bench and BrowseComp benchmarks. On SWE-Bench Verified, GCC improves task resolution by over 13\% relative to strong long-context baselines and outperforms 26 existing open and commercial systems, reaching over 80\% success rate. The project will be open-sourced for the research community.

Packed high-token 3D fields are not just long-sequence modeling: once a volume is packed into tokens, sequence order is no longer the geometry of the sample. This creates a general failure mode for attention over packed geometric data: sequence-index positional mechanisms can learn the memory layout rather than the object. We introduce Pack3D, a geometry-aware local--global attention operator that separates memory layout from attention geometry. Pack3D keeps local neighborhood attention exact, compresses only distant context into contiguous 3D block summaries, and indexes attention with axial RoPE on true 3D token-grid coordinates rather than packed sequence positions. With RoPE applied before block compression, each query interacts with a geometry-aware mixture of block tokens, so compressed global context remains spatially interpretable after packing. A synthetic repacking benchmark makes the failure mode explicit: learned positional embeddings and packed-sequence RoPE fail under layout changes, whereas axial grid-coordinate RoPE remains invariant. On Allen WTC--11 hiPSC microscopy, our main claim-grade domain, Pack3D improves matched learned-position local--global attention by 0.0978 best macro Dice and 0.1248 final macro Dice at 262k tokens and depth 8. In an Allen-15 joint comparison, Pack3D also improves over a parameter-matched TokenGrid 3D U-Net on both best and final macro Dice. Inference-time ablations show that the trained model genuinely uses its compressed summaries, and the fused forward/backward implementation is $6.49\times$ lower latency than dense SDPA at 65k tokens.

Error Feedback (EF) has become a core mechanism for aggressive gradient compression in communication-efficient distributed training, and is now used in geo-distributed frameworks such as DiLoCo. While existing theory mainly explains its optimization convergence, the generalization behavior of EF remains poorly understood, especially under nonlinear biased compression and distributed local updates. We present the first generalization analysis of EF for both single-node EF-SGD and DiLoCo-EF through an optimization-driven on-average model stability framework, without imposing bounded-gradient assumptions. Technically, we develop a buffer-aware Lyapunov argument that tracks the coupled dynamics of the model and error buffer, while co-coercivity absorbs same-sample gradient-difference energy into the optimization trajectory. For convex losses, our bounds show that compression-induced stability terms are transient, decaying as $\mathcal{O}(\sqrt{b_\alpha}\,T^{-3/4}+b_\alpha T^{-1})$, and therefore do not create a non-vanishing generalization floor beyond the dominant $\mathcal{O}(T^{-1/2})$ optimization term. In the DiLoCo setting, worker averaging further suppresses compression perturbations, and a local stepsize $\eta_l=\Theta(1/(H\sqrt{T}))$ controls the drift from $H$ local steps, yielding the leading $\mathcal{O}(1/\sqrt{MHT})$ excess-risk scaling. Experiments on convex benchmarks corroborate the theory, showing that Top-$k$, Random-$k$, and quantization with matched contraction factor $\alpha$ exhibit nearly indistinguishable generalization trajectories.

Long-context DNA models are limited by token-mixing cost and by how compression allocates representational budget across the genome. Existing approaches operate close to base-pair resolution, apply fixed downsampling, or learn content-dependent chunks without an explicit genomic budget, making long-context pretraining expensive and difficult to control. We introduce GeneZip, a region-aware DNA compression framework that combines H-Net-style dynamic routing with a Region-Aware Ratio (RAR) objective and bounded routing. GeneZip uses static gene-structure annotations during compression training to specify region-wise base-pairs-per-token (BPT) targets; at inference time, it compresses raw unseen DNA without annotations. GeneZip provides three main benefits. First, it is effective: GeneZip variants achieve the best validation PPL among encoder-based compressors, with GeneZip-70M already operating at 137.6 BPT, and across four reproducible DNALongBench tasks---contact map prediction, eQTL prediction, enhancer--target gene prediction, and transcription-initiation signal prediction---GeneZip obtains the best average rank among compared sequence models. Second, it is redundancy-aware: a post-hoc RepeatMasker/TRF analysis shows that, without repeat supervision, GeneZip assigns higher local BPT to TE-derived interspersed repeats and tandem repeats, two major classes of repetitive DNA sequence redundancy. Third, it is efficient: by reducing the effective token-mixing length, GeneZip enables longer-context and larger-capacity pretraining, including 128K-context and 636M-parameter variants on a single A100 80GB GPU, and fine-tunes the eQTL task 50.4$\times$ faster than JanusDNA (50 vs. 2520 minutes). These results establish GeneZip as an effective, redundancy-aware, and efficient compression interface for long-context DNA modeling.

Immersive video communication requires photorealistic, render-efficient, and compact dynamic scene representations. 3D Gaussian Splatting (3DGS) offers a promising representation, but dynamic 3DGS remains difficult to compress due to dense primitives and spatiotemporal redundancy. Anchor-based formulations improve compactness with sparse scaffolds that share geometry and appearance across primitives. However, existing designs often rely on deforming a single canonical scaffold and condition each primitive on its associated anchor in isolation, limiting their ability to handle non-local dynamics and disocclusion while under-exploiting inter-anchor correlations, particularly in motion- or texture-dense regions. To address these limitations, we propose SAGA, a volumetric video codec built upon Sparse Anchor-assisted GAussian splatting representations. SAGA represents dynamic 3D scenes using hierarchically organized sparse 4D anchors, where coordinate-based INR decoders generate fine anchors and Gaussian primitives from inter-anchor interpolations, enabling compact parameter sharing across spatiotemporal structures. For long-range dependencies among unstructured anchors, we further introduce fixed-size memory slots with orthogonality-informed updates for accurate entropy-context modeling. Experiments show SAGA outperforms the state-of-the-art 3DGS-based codec GIFStream, achieving BD-rate reductions of 84.54% on Neu3D, 76.71% on Panoptic Sports, and 83.94% on MPEG MIV.

In real-world tasks such as robotic control and autonomous driving, agents often face non-Markovian environments due to incomplete observations. Solving such information-impoverished decision tasks requires reconstructing the context from history. However, existing recurrent encoders have two fundamental limitations: First, the model suffers from a Markovian compression bottleneck, i.e., it only processes observation history as input but neglects thought process history. This leads to the loss of deep semantic cues from early interactions, which are easily overwritten by noise over time due to recursion. Second, it lacks the adaptive capability to adjust the memory span for different tasks. To address these limitations, we propose FracEncoder, a context encoder based on fractional-order neural differential equations. First, FracEncoder leverages the nonlocal property of the fractional-order derivative and uses a power-law kernel to directly incorporate the trajectory of the thought process into the evolution of the current state, effectively overcoming the temporal locality limitations of traditional models. Second, the fractional order $\alpha$ is set as an end-to-end learnable parameter. It explicitly characterizes how strongly the system anchors to its history, so the model can discover a suitable memory span for each task. We evaluate FracEncoder on more than ten benchmarks across three settings, namely partially observable tasks, meta reinforcement learning, and delayed-observation control. FracEncoder consistently matches or outperforms representative baselines, including GPIDE, GRU-ODE, ReSeL, and LRU. On the challenging 8-step delayed-observation task, it reaches average returns of $-511.5 \pm 7.9$ on Pendulum and $732.1 \pm 152.7$ on Hopper. As a plug-in module, FracEncoder also yields stable gains across different reinforcement learning algorithms. Ablation studies further show that the learned $\alpha$ aligns with the intrinsic memory demand of each task, providing an interpretable physical measure of the cognitive complexity of non-Markovian tasks. The code is available at https://anonymous.4open.science/r/FracEncoder-60BA.

Long-context LLM agents often struggle with growing token, memory, and latency costs, making efficient context compression essential for practical deployment. Existing LLM-as-a-compressor methods remain noticeably inferior to using the full context. We find that this gap partly stems from their inability to preserve contextual information effectively. In this work, we revisit context compression from a structural perspective and identify two key bottlenecks in standard LLM-based compressors: limited coordination among compression tokens during information aggregation, and layerwise dilution that weakens useful signals from intermediate hidden states. To address these limitations, we propose **ComprExIT** (Context **Compr**ession via **Ex**plicit **I**nformation **T**ransmission), a new context compression framework based on explicit information transmission. ComprExIT adaptively selects features across frozen LLM layers, then allocates information from anchors to compression slots through a globally coordinated transport plan. Experiments on 12 datasets show that ComprExIT consistently outperforms strong soft-compression baselines, improving average F1 by up to 18.5%, while adding only ~1% trainable parameters and achieving more than 2× faster compression than the fastest baselines. The code will be released upon acceptance.

The quadratic computational complexity of attention during prefilling remains a primary bottleneck for the efficiency of long-context LLMS. While sparse attention methods mitigate this cost, existing dynamic approaches rely on computationally expensive $O(N^2)$ scoring, which undermines their practical efficiency gains. In this paper, we observe that long-context attention matrices frequently exhibit prominent diagonal stripe patterns, indicating that query-key pairs separated by specific relative distances yield consistently high correlation. Motivated by this, we propose FFTSparse, a hybrid block-sparse attention framework. By formulating the identification of these dominant attention lags as cross-correlation in the frequency domain via the Fast Fourier Transform (FFT), FFTSparse reduces the cost to identify the sparsity pattern of diagonal stripes to $O(N \log N)$. Additionally, we introduce a lightweight identification module that categorizes the patterns of attention heads and then applies tailored masking strategies accordingly. Extensive evaluations demonstrate that FFTSparse incurs a negligible mask construction overhead of merely 0.02\% relative to full attention. Furthermore, it maintains competitive downstream performance on long-context benchmarks while achieving up to a $7.4x$ prefilling speedup at a 128K sequence length, offering a highly efficient solution for long-context inference of LLMS.

Transformers predict over a representation of a sequence. The same data can be written as bytes, characters, or subword tokens, and these representations may be lossless. Yet, under a fixed context window, they need not expose the same information to the model. This raises a basic question: how does the choice of representation change what a finite-context predictor can achieve? We study this question on Markov sources and uncover two complementary phenomena. First, we observe that moving to smaller representation units can hurt prediction even when the context window is enlarged to cover the relevant source history. To explain this, we introduce *fragmentation*: a lossless recoding that replaces each source symbol by several smaller units. We prove that fragmentation can strictly increase the optimal finite-context log-loss, showing that the gap is not merely an optimization or capacity issue, but can be intrinsic to the representation. Second, we study the opposite direction: greedy tokenization, which groups source symbols into larger units. We show that tokenization can make a short token window behave like a longer source-context window, and we give a loss guarantee describing when this is achievable. The guarantee depends on how reliably token windows span the needed source history, together with the compression rate of the tokenizer. This also yields a simple diagnostic for real tokenizers: measuring how much source context a fixed token window reliably contains. Together, the two directions give a finite-context view of why representation matters for Transformer prediction.

A central question in modern learning theory is what notion of compression, if any, explains why over-parameterized neural networks generalize. Existing answers measure complexity at different levels, ranging from weights to representations, and estimating the associated quantities either requires held-out data in principle or relies on mutual-information estimation in ultra-high dimensions. We approach the question through the minimum description length (MDL) principle and propose the prequential regret as a complementary signal that is computable from the training trajectory alone. We establish that, in the (nearly) interpolating regime, the expected prequential regret upper-bounds the description length of the trained parameter under the algorithm-induced distribution, and that this description length translates into a high-probability generalization bound in the PAC-MDL setting. Along the way, we show that the encoder-complexity correction recently added to the algorithm-level information bottleneck is upper-bounded by an MDL parameter description length, bringing the two views of compression onto a shared footing. Empirically, the prequential regret tracks the generalization error on CIFAR-10 and CIFAR-100, with Pearson correlations exceeding 0.9.

Long-context LLM inference is fundamentally a retrieval process, transitioning from diffuse to sharply focused attention at a few architecture-determined (CTLs). Existing acceleration methods leave significant costs unaddressed: sparse attention ignores FFNs, KV cache compression neglects prefill, and hidden-state pruning relies on fixed, heuristic rules. We introduce CutAttn, a training-free framework that exploits CTLs via inter-layer Jensen--Shannon divergence, pruning redundant context at chunk granularity while dynamically gated by a normalized-entropy criterion. Operating at the hidden-state level, CutAttn reduces both attention and FFN costs, composing orthogonally with existing acceleration techniques. Evaluated on RULER, InfiniteBench, and LongBench, CutAttn preserves full-attention accuracy while achieving up to a $2.02\times$ prefill speedup on 128K sequences (and $2.81\times$ when composed with sparse attention). Furthermore, our theoretical analysis provides approximation bounds, an entropy-compression limit justifying the gating, and a closed-form speedup ratio.

Pipeline-parallel distributed optimization is essential for large-scale machine learning but is challenged by significant communication overhead from transmitting high-dimensional activations and gradients between workers. Existing approaches often depend on impractical unbiased gradient assumptions or incur sample-size memory overhead. This paper introduces Clapping, a Communication compression algorithm with LAzy samPling for Pipeline-parallel learnING. Clapping adopts a lazy sampling strategy that reuses data samples across steps, breaking sample-wise memory barrier and supporting convergence in few-epoch or online regimes. Clapping comprises two variants including Clapping-FC and Clapping-FU, both of which achieve convergence without unbiased assumption for compressed gradient, effectively addressing compression error propagation in multi-worker settings. Numerical experiments validate the performance of Clapping across different tasks.