arXiv subjects

Adam Deng

Publications and source records attributed to Adam Deng.

5 recordsLinked to original sources

arXiv2026

ParaDeAl: Parallel Descent Algorithms for Sparse Statistics

Best subset selection is a central problem in statistical regression, drawing increasing attention from the broader optimization and machine learning communities. Solving the hard L0 sparsity constraint directly has been computationally challenging; as a result, most existing work proposes formulations which do not enforce the constraint, but instead use a mix of L0/L1/L2 penalties to encourage sparsity. Modern methods for solving such penalized formulations are tractable for p = 100,000, but give up direct control over sparsity. In this paper, we propose a two-stage proximal gradient descent-based approach which both enforces the sparsity constraint and incorporates L1/L2 penalties, leveraging GPUs for computational acceleration to find approximate solutions. Our method, ParaDeAl, offers the user the ability to select the problem hyperparameters, solves all configurations in parallel, and obtains solutions up to 15x faster than the best state-of-the-art CPU competitors while achieving almost identical test accuracy for the same support size for large (p > 100,000) synthetic and real-world datasets.

math.OC↗

arXiv2026

Close Enough is Nowhere Near: Fidelity-Consumability Decoupling in KV Caches

Multi-agent LLM systems communicate in text, which is lossy and wasteful: the sender throws away the reading it already did. A growing literature instead transmits the KV cache, the natural carrier of that reading. We study the most appealing form of the idea: both models stay frozen and the sender's cache is substituted for the reader's with nothing trained. It does not work, and the obstruction belongs to the reader. A frozen transformer reads only caches close to the one it would have written itself, a region we call its tolerable set. We show that geometric similarity to that cache does not predict whether the reader can use it, and the failure is not gradual, so a more faithful substitute is not the remedy it appears to be. We explain this phenomenon with competitive attention. Experimentally, we test eight diverse substitution methods against two measured anchors: an uninformative cache baseline and the reader's own accuracy from having read the passage. All sit at the baseline. That turns a long tail of separate negative results into one outcome with a stated condition for overturning it. A different design does work: if the reader keeps its own cache and attends over the sender's as extra context, adapting only its queries, it beats its own baseline, and also clears the baseline with eight senders pooled where undirected attention over the same union falls below it. We offer this as a characterization on one probe, not a leaderboard claim. Our central contribution is a redirection: pursuing faithful substitute caches is the wrong target; adding a cache to what the reader already has is the promising one.

cs.LG↗

arXiv2026

Augment the Cache, Replace the Activation: A 2x2 Study of Internal-State Sharing

Multi-agent systems built from frozen language models communicate through text, which discards the internal state the sender already computed and makes the receiver pay to rebuild it. Work on internal channels avoids that cost, framing the problem as transfer: agent A sends information to agent B. Every agent pair that must communicate needs its own transfer bridge, and every agent added to a system needs a new bridge fitted against each previous agent. We reframe this problem as sharing: an agent publishes its internal state once into a pool, and every peer reads that pool through a single mechanism fitted to no sender in particular, so adding an agent adds a publisher and no new bridge. Whereas the carrier of the internal state and the fusion mechanism are usually chosen independently in the literature, our main finding is that these two choices are coupled; the deciding factor is where the carrier enters the receiver's computation. Consumability is a property of the receiving computation rather than of the representation. We find that the key-value (KV) cache must be augmented rather than swapped in: the receiver cross-attends over a peer's cache while keeping its own, and with only a small query-side adapter trained, matches the accuracy it gets from reading the source document. Activations invert this. They are best written straight into the receiver, with no bridge, no adapter and no training, which beats exposing them alongside the receiver's own, though that advantage rests on knowing where in the receiver the state belongs. The rule is augment the KV cache, replace the activation, established under one protocol with a shared control ladder. One adapter, trained once, also serves publishers it never trained on, and a pool of several at once, but only those whose internal state lies close to the receiver's own.

cs.LG↗

arXiv2026

ADMM-Q: An Improved Hessian-based Weight Quantizer for Post-Training Quantization of Large Language Models

Quantization is an effective strategy to reduce the storage and computation footprint of large language models (LLMs). Post-training quantization (PTQ) is a leading approach for compressing LLMs. Popular weight quantization procedures, including GPTQ and RTN, suffer in model utility, especially at aggressive quantization levels (sub-4-bit). We propose ADMM-Q, a novel weight quantization algorithm that considers the layer-wise quantization problem. Our algorithm is based on a combinatorial variant of the Alternating Direction Method of Multipliers (ADMM). Our operator-splitting procedure updates weights continuously to minimize the layer-wise reconstruction error, while gradually enforcing the quantization constraints with convergence guarantees. We propose additional algorithmic enhancements (e.g., penalty scheduling, preconditioning, and a local search post-processing step) to make ADMM-Q efficient at LLM scale. ADMM-Q is modular and can be used as a drop-in replacement for any weight quantizer within existing quantization pipelines: ADMM-Q is fully composable with existing techniques including range clipping, learned or random rotations, and activation scaling. Using ADMM-Q in place of GPTQ on Qwen3-8B, we decrease WikiText-2 perplexity in: (i) the W3A16 weight-only setting (12.85 $\rightarrow$ 10.06); (ii) the W4A8 SmoothQuant procedure (9.29 $\rightarrow$ 8.68); and (iii) the W2A4KV4 SpinQuant procedure (66.11 $\rightarrow$ 19.42).

cs.LG↗

arXiv2023

Coordinating mobile network coverage and vehicle routing: a double column generation approach

The emergence of 5G technologies opens opportunities to support mission-critical activities with high-speed Internet coverage. This paper defines a joint job-emitting vehicle routing problem with time windows to coordinate the operations of mission-oriented vehicles ("mission vehicles") and mobile emitters ("emitting vehicles"). This problem exhibits a joint vehicle routing structure, with coupling constraints to ensure that each job is supported by appropriate network coverage. We solve it via an exact and finite double column generation algorithm: pricing problems generate vehicle paths dynamically, and a master problem coordinates the operations of mission vehicles and emitting vehicles to ensure appropriate network coverage for each job. We propose several acceleration strategies to strengthen the algorithm's computational performance. Computational results show the scalability of the proposed methodology. Specifically, the methodology cuts runtimes by over 95% in small-scale instances as compared to an explicit formulation, and scales to large-scale instances involving over 50 jobs. From a practical standpoint, results highlight the benefits of dynamically coordinating mission vehicles and emitting vehicles, thus suggesting opportunities to support emerging 5G technologies with dedicated routing algorithms.

math.OC↗