Chris Lott

Raghavv Goel, Mukul Gagrani, Mingu Lee et al.

Large language models (LLMs) exhibit latent multi-token prediction (MTP) capabilities despite being trained solely for next-token generation. We propose a simple, training-free MTP approach that probes an LLM using on-the-fly mask tokens drawn from its embedding space, enabling parallel prediction of future tokens without modifying model weights or relying on auxiliary draft models. Our method constructs a speculative token tree by sampling top-K candidates from mask-token logits and applies a lightweight pruning strategy to retain high-probability continuations. During decoding, candidate predictions are verified in parallel, resulting in lossless generation while substantially reducing the number of model calls and improving token throughput. Across benchmarks, our probing-based MTP consistently outperforms existing training-free baselines, increasing acceptance length by approximately 12\% on LLaMA3 and 8--12\% on Qwen3, and achieving throughput gains of up to 15--19\%. Finally, we provide theoretical insights and empirical evidence showing that decoder layers naturally align mask-token representations with next-token states, enabling accurate multi-step prediction without retraining or auxiliary models.

Dalton Jones, Junyoung Park, Matthew Morse et al.

We present QUOKA: Query-oriented KV selection for efficient attention, a training-free and hardware agnostic sparse attention algorithm for accelerating transformer inference under chunked prefill. While many queries focus on a smaller group of keys in the attention operator, we observe that queries with low cosine similarity with respect to the mean query interact more strongly with more keys and have the greatest contribution to final attention logits. By prioritizing these low cosine similarity queries, the behavior of full attention during the prefill stage can be closely approximated. QUOKA leverages this observation, accelerating attention by (1) first retaining a small set of representative queries and (2) then subselectin the keys most aligned with those queries. Through experiments on Needle-In-A-Haystack, LongBench, RULER, and Math500, we show that, while realizing a 3x reduction in time-to-first-token, 5x speedup in attention on Nvidia GPUs and up to nearly a 7x speedup on Intel Xeon CPUs, QUOKA achieves near-baseline accuracy, utilizing 88% fewer key-value pairs per attention evaluation.

Junyoung Park, Dalton Jones, Matthew J Morse et al.

We demonstrate that geometrically distinctive keys during LLM inference tend to have high attention scores. Based on the phenomenon we propose KeyDiff, a training-free KV cache eviction method based solely on key similarity. Unlike other KV cache eviction methods, KeyDiff can process arbitrarily long prompts within strict resource constraints and efficiently generate responses. We provide a theoretical basis for KeyDiff by relating key diversity with attention scores. These results imply KeyDiff can efficiently identify the most important tokens to retain. Notably KeyDiff does not rely on attention scores, allowing the use of optimized attention mechanisms like FlashAttention. Under a strict memory allowance, we demonstrate the effectiveness of KeyDiff for the Llama and Qwen model families by observing a performance gap of less than 0.04% with 8K cache budget ($\sim$23% KV cache reduction) from the non-evicting baseline on LongBench for Llama 3.1-8B and Llama 3.2-3B. We also observe near baseline performance for Deepseek-R1-Distill-Llama-8B on the Math500 reasoning benchmark and decrease end-to-end inference latency by up to 30% compared to the other token-eviction methods.

Raghavv Goel, Risheek Garrepalli, Sudhanshu Agrawal et al.

Autoregressive (AR) language models form representations incrementally through left-to-right prediction, whereas diffusion language models (dLLMs) are trained via full-sequence denoising. Although recent dLLMs match AR performance, it remains unclear whether diffusion objectives fundamentally reshape internal representations across depth. We perform the first layer- and token-wise representational analysis comparing native dLLMs (LLaDA), native AR models (Qwen2.5), and AR-initialized dLLMs (Dream-7B). We find that diffusion objectives result in different, more hierarchical abstractions with substantial early-layer redundancy and reduced recency bias, while AR objectives produce tightly coupled, depth-dependent representations. Critically, AR-initialized dLLMs retain AR-like representational dynamics despite diffusion training, revealing persistent initialization bias. Leveraging this observed representational redundancy, we introduce a static, task-agnostic inference-time layer-skipping method requiring no architectural changes or KV-cache sharing. Native dLLMs achieve up to 18.75% FLOPs reduction while preserving over 90% performance on reasoning and code generation benchmarks, whereas AR models degrade sharply under comparable skipping. These results link training objectives to representational structure and enable practical, cache-orthogonal efficiency gains.

Raghavv Goel, Sudhanshu Agrawal, Mukul Gagrani et al.

In this paper, we introduce a simple training-free technique to improve the performance of drafter-based speculative decoding (SpD) methods that incorporates language modeling head (LM head) during drafting process. A drafter-based speculative decoding leverages one or more smaller language models, a.k.a. drafters or draft models, to sample a draft sequence or tree consisting of multiple tokens, followed by verification by a base LLM, a target model, accepting a subset as its valid generation. As it is usually considered that the speculative decoding requires one-to-one mapping between vocabularies of the target model and the draft model, it has been natural to share the vocabulary between them, or even share the LM head as in EAGLE or Medusa. We first identify that this draft token sampling scheme inherently contains an unnecessary inference overhead in drafting, especially for some target LLMs with very large vocabularies. Then, we propose a simple technique, VocabTrim, to mitigate the drafting overhead to improve the generation speed in memory-bound environment. VocabTrim reconstructs the drafter LM head to contain only a limited set of tokens, selected by the most frequently sampled from the vocabulary of the target model. While limiting the vocabulary in drafting slightly degrades the acceptance rate, it significantly reduces the drafting latency in memory-bound process which is often the case on edge devices, resulting in higher memory-bound speed up (MBSU). We show that our method can boost the memory-bound speed-up for Llama-3 models on Spec-Bench, specifically by 16% for Llama-3.2-3B-Instruct.

Raghavv Goel, Junyoung Park, Mukul Gagrani et al.

While long context support of large language models has extended their abilities, it also incurs challenges in memory and compute which becomes crucial bottlenecks in resource-restricted devices. Token eviction, a widely adopted post-training methodology designed to alleviate the bottlenecks by evicting less important tokens from the cache, typically uses attention scores as proxy metrics for token importance. However, one major limitation of attention score as a token-wise importance metrics is that it lacks the information about contribution of tokens to the attention output. In this paper, we propose a simple eviction criterion based on the contribution of cached tokens to attention outputs. Our method, CAOTE, optimizes for eviction error due to token eviction, by seamlessly integrating attention scores and value vectors. This is the first method which uses value tokens on top of attention-based eviction scores in closed-form. Additionally, CAOTE can act as a meta-heuristic method with flexible usage with any token eviction method. We show that CAOTE, when combined with the state-of-the-art attention score-based methods, always improves accuracies on the downstream task, indicating the importance of leveraging information from values during token eviction process.

Bert Moons, Parham Noorzad, Andrii Skliar et al.

Current state-of-the-art Neural Architecture Search (NAS) methods neither efficiently scale to multiple hardware platforms, nor handle diverse architectural search-spaces. To remedy this, we present DONNA (Distilling Optimal Neural Network Architectures), a novel pipeline for rapid, scalable and diverse NAS, that scales to many user scenarios. DONNA consists of three phases. First, an accuracy predictor is built using blockwise knowledge distillation from a reference model. This predictor enables searching across diverse networks with varying macro-architectural parameters such as layer types and attention mechanisms, as well as across micro-architectural parameters such as block repeats and expansion rates. Second, a rapid evolutionary search finds a set of pareto-optimal architectures for any scenario using the accuracy predictor and on-device measurements. Third, optimal models are quickly finetuned to training-from-scratch accuracy. DONNA is up to 100x faster than MNasNet in finding state-of-the-art architectures on-device. Classifying ImageNet, DONNA architectures are 20% faster than EfficientNet-B0 and MobileNetV2 on a Nvidia V100 GPU and 10% faster with 0.5% higher accuracy than MobileNetV2-1.4x on a Samsung S20 smartphone. In addition to NAS, DONNA is used for search-space extension and exploration, as well as hardware-aware model compression.

Yang Yang, Anusha Lalitha, Jinwon Lee et al.

This paper proposes a novel pipeline for automatic grammar augmentation that provides a significant improvement in the voice command recognition accuracy for systems with small footprint acoustic model (AM). The improvement is achieved by augmenting the user-defined voice command set, also called grammar set, with alternate grammar expressions. For a given grammar set, a set of potential grammar expressions (candidate set) for augmentation is constructed from an AM-specific statistical pronunciation dictionary that captures the consistent patterns and errors in the decoding of AM induced by variations in pronunciation, pitch, tempo, accent, ambiguous spellings, and noise conditions. Using this candidate set, greedy optimization based and cross-entropy-method (CEM) based algorithms are considered to search for an augmented grammar set with improved recognition accuracy utilizing a command-specific dataset. Our experiments show that the proposed pipeline along with algorithms considered in this paper significantly reduce the mis-detection and mis-classification rate without increasing the false-alarm rate. Experiments also demonstrate the consistent superior performance of CEM method over greedy-based algorithms.