Hasil untuk "Speculative philosophy"

Elia Peruzzo, Guillaume Sautière, Amirhossein Habibian

Autoregressive (AR) models have achieved remarkable success in image synthesis, yet their sequential nature imposes significant latency constraints. Speculative Decoding offers a promising avenue for acceleration, but existing approaches are limited by token-level ambiguity and lack of spatial awareness. In this work, we introduce Multi-Scale Local Speculative Decoding (MuLo-SD), a novel framework that combines multi-resolution drafting with spatially informed verification to accelerate AR image generation. Our method leverages a low-resolution drafter paired with learned up-samplers to propose candidate image tokens, which are then verified in parallel by a high-resolution target model. Crucially, we incorporate a local rejection and resampling mechanism, enabling efficient correction of draft errors by focusing on spatial neighborhoods rather than raster-scan resampling after the first rejection. We demonstrate that MuLo-SD achieves substantial speedups - up to $\mathbf{1.7\times}$ - outperforming strong speculative decoding baselines such as EAGLE-2 and LANTERN in terms of acceleration, while maintaining comparable semantic alignment and perceptual quality. These results are validated using GenEval, DPG-Bench, and FID/HPSv2 on the MS-COCO 5k validation split. Extensive ablations highlight the impact of up-sampling design, probability pooling, and local rejection and resampling with neighborhood expansion. Our approach sets a new state-of-the-art in speculative decoding for image synthesis, bridging the gap between efficiency and fidelity.

Jaydip Sen, Subhasis Dasgupta, Hetvi Waghela

Speculative decoding has emerged as an effective approach for accelerating autoregressive inference by parallelizing token generation through a draft-then-verify paradigm. However, existing methods rely on static drafting lengths and rigid verification criteria, limiting their adaptability across varying model uncertainties and input complexities. This paper proposes an information-theoretic framework for speculative decoding based on confidence-modulated drafting. By leveraging entropy and margin-based uncertainty measures over the drafter's output distribution, the proposed method dynamically adjusts the number of speculatively generated tokens at each iteration. This adaptive mechanism reduces rollback frequency, improves resource utilization, and maintains output fidelity. Additionally, the verification process is modulated using the same confidence signals, enabling more flexible acceptance of drafted tokens without sacrificing generation quality. Experiments on machine translation and summarization tasks demonstrate significant speedups over standard speculative decoding while preserving or improving BLEU and ROUGE scores. The proposed approach offers a principled, plug-in method for efficient and robust decoding in large language models under varying conditions of uncertainty.

Yudi Zhang, Weilin Zhao, Xu Han et al.

Speculative decoding and quantization effectively accelerate memory-bound inference of large language models. Speculative decoding mitigates the memory bandwidth bottleneck by verifying multiple tokens within a single forward pass, which increases computational effort. Quantization achieves this optimization by compressing weights and activations into lower bit-widths and also reduces computations via low-bit matrix multiplications. To further leverage their strengths, we investigate the integration of these two techniques. Surprisingly, experiments applying the advanced speculative decoding method EAGLE-2 to various quantized models reveal that the memory benefits from 4-bit weight quantization are diminished by the computational load from speculative decoding. Specifically, verifying a tree-style draft incurs significantly more time overhead than a single-token forward pass on 4-bit weight quantized models. This finding led to our new speculative decoding design: a hierarchical framework that employs a small model as an intermediate stage to turn tree-style drafts into sequence drafts, leveraging the memory access benefits of the target quantized model. Experimental results show that our hierarchical approach achieves a 2.78$\times$ speedup across various tasks for the 4-bit weight Llama-3-70B model on an A100 GPU, outperforming EAGLE-2 by 1.31$\times$. Code available at https://github.com/AI9Stars/SpecMQuant.

Yiwei Li, Jiayi Shi, Shaoxiong Feng et al.

We propose a novel speculative decoding method tailored for multi-sample reasoning scenarios, such as self-consistency and Best-of-N sampling. Our method exploits the intrinsic consensus of parallel generation paths to synthesize high-quality draft tokens without requiring auxiliary models or external databases. By dynamically analyzing structural patterns across parallel reasoning paths through a probabilistic aggregation mechanism, it identifies consensus token sequences that align with the decoding distribution. Evaluations on mathematical reasoning benchmarks demonstrate a substantial improvement in draft acceptance rates over baselines, while reducing the latency in draft token construction. This work establishes a paradigm shift for efficient multi-sample inference, enabling seamless integration of speculative decoding with sampling-based reasoning techniques.

Michele Lobo

This commentary illuminates how Whitehead's vitalistic ethos and speculative philosophy mobilises decolonial leaps in more-than-human geographies. These risky leaps that unsettle apocalyptic, commonsense western literacies of planetary crises call for daring and experimentation. Amid the ongoing brutality of a racial, colonial, and capitalist logics, perhaps Whitehead and Roberts are accomplices in decolonial leaps that contribute to a planetary consciousness.

Oscar Brown, Zhengjie Wang, Andrea Do et al.

The acceleration of Large Language Models (LLMs) with speculative decoding provides a significant runtime improvement without any loss of accuracy. Currently, EAGLE-2 is the state-of-the-art speculative decoding method, improving on EAGLE with a dynamic draft tree. We introduce Dynamic Depth Decoding (DDD), which optimises EAGLE-2's tree drafting method using a dynamic depth. This extends the average speedup that EAGLE-2 achieves over EAGLE by $44\%$, giving DDD an average speedup of $3.16$x.

Heming Xia, Zhe Yang, Qingxiu Dong et al.

To mitigate the high inference latency stemming from autoregressive decoding in Large Language Models (LLMs), Speculative Decoding has emerged as a novel decoding paradigm for LLM inference. In each decoding step, this method first drafts several future tokens efficiently and then verifies them in parallel. Unlike autoregressive decoding, Speculative Decoding facilitates the simultaneous decoding of multiple tokens per step, thereby accelerating inference. This paper presents a comprehensive overview and analysis of this promising decoding paradigm. We begin by providing a formal definition and formulation of Speculative Decoding. Then, we organize in-depth discussions on its key facets, such as drafter selection and verification strategies. Furthermore, we present a comparative analysis of leading methods under third-party testing environments. We aim for this work to serve as a catalyst for further research on Speculative Decoding, ultimately contributing to more efficient LLM inference.

Doohyuk Jang, Sihwan Park, June Yong Yang et al.

Auto-Regressive (AR) models have recently gained prominence in image generation, often matching or even surpassing the performance of diffusion models. However, one major limitation of AR models is their sequential nature, which processes tokens one at a time, slowing down generation compared to models like GANs or diffusion-based methods that operate more efficiently. While speculative decoding has proven effective for accelerating LLMs by generating multiple tokens in a single forward, its application in visual AR models remains largely unexplored. In this work, we identify a challenge in this setting, which we term \textit{token selection ambiguity}, wherein visual AR models frequently assign uniformly low probabilities to tokens, hampering the performance of speculative decoding. To overcome this challenge, we propose a relaxed acceptance condition referred to as LANTERN that leverages the interchangeability of tokens in latent space. This relaxation restores the effectiveness of speculative decoding in visual AR models by enabling more flexible use of candidate tokens that would otherwise be prematurely rejected. Furthermore, by incorporating a total variation distance bound, we ensure that these speed gains are achieved without significantly compromising image quality or semantic coherence. Experimental results demonstrate the efficacy of our method in providing a substantial speed-up over speculative decoding. In specific, compared to a naïve application of the state-of-the-art speculative decoding, LANTERN increases speed-ups by $\mathbf{1.75}\times$ and $\mathbf{1.82}\times$, as compared to greedy decoding and random sampling, respectively, when applied to LlamaGen, a contemporary visual AR model. The code is publicly available at https://github.com/jadohu/LANTERN.

Xiaoxuan Liu, Jongseok Park, Langxiang Hu et al.

Large Language Model (LLM) serving systems batch concurrent user requests to achieve efficient serving. However, in real-world deployments, such inter-request parallelism from batching is often limited by external factors such as low request rates or memory constraints. Recent works focus on intra-request parallelism from speculative decoding as a solution to this problem. Unfortunately, benefits from intra-request parallelism are often fragile, as speculative decoding causes overhead, and speculated tokens may miss. We observe that speculative decoding may degrade LLM serving performance if added naively without tuning to the incoming requests and the speculation method. To alleviate the need for expert tuning and make speculative decoding more robust, we present TurboSpec, a speculation control system that automatically profiles the execution environment and utilizes a feedback-based algorithm to dynamically adjust the amount of intra-request parallelism in LLM serving. TurboSpec predicts "goodput" - the amount of successfully generated tokens - to evaluate and adjust intra-request parallelism amount to that with the highest goodput in runtime. We implement TurboSpec on a real-world LLM serving system vLLM and demonstrate its effectiveness across diverse workloads and hardware configurations, providing consistent performance improvements across all test scenarios.

Zili Wang, Robert Zhang, Kun Ding et al.

Continuous visual autoregressive (AR) models have demonstrated promising performance in image generation. However, the heavy autoregressive inference burden imposes significant overhead. In Large Language Models (LLMs), speculative decoding has effectively accelerated discrete autoregressive inference. However, the absence of an analogous theory for continuous distributions precludes its use in accelerating continuous AR models. To fill this gap, this work presents continuous speculative decoding, and addresses challenges from: 1) low acceptance rate, caused by inconsistent output distribution between target and draft models, and 2) modified distribution without analytic expression, caused by complex integral. To address challenge 1), we propose denoising trajectory alignment and token pre-filling strategies. To address challenge 2), we introduce acceptance-rejection sampling algorithm with an appropriate upper bound, thereby avoiding explicitly calculating the integral. Furthermore, our denoising trajectory alignment is also reused in acceptance-rejection sampling, effectively avoiding repetitive diffusion model inference. Extensive experiments demonstrate that our proposed continuous speculative decoding achieves over $2\times$ speedup on off-the-shelf models, while maintaining the original generation quality. Codes is available at: https://github.com/MarkXCloud/CSpD

Ігор ІЩЕНКО

The first chapter reveals the international aspect of the synergy of religion and politics in the digital age. The author analyzed the specifics of various manifestations of religious beliefs in the environment against the backdrop of extreme events associated with revolutions and the COVID-2019 pandemic. Digital platforms serve as regulators of the religious impulse, which can both stabilize the social situation and transfer it to a state of bifurcation. The author paid considerable attention to the study of the impact of the digitalization of religion on young people. We are exploring the problem of how religious digital practices today find broad support from the international community, how they manifest themselves at the everyday level, and cause the formation of new traditions and rituals based on online myth-making.

R. Solomon, K. Higgins

1. From Leibniz to KantLewis White Beck, University of Rochester2. Kant's copernican revolutionDaniel Bonevac, University of Texas at Austin3. Kant's moral philosophyDon Becker, University of Texas at Austin4. Kant: Critique of JudgmentPatrick Gardiner, Magdalen College, Oxford 5. Fichte and Schelling: the jena periodDan Breazeale, University of Kentucky6. Hegel's Phenomenology of SpiritRobert C.Solomon, University of Texas at Austin7. Hegel's logic and philosophy of mindWillem deVries, University of New Hampshire8. Hegel, spirit, and politicsLeo Rauch, Babson College9. The young Hegelians, Feuerbach and MarxRobert Nola, University of Auckland10. Arthur SchopenhauerKathleen M.Higgins, University of Texas at Austin11. Kierkegaard's speculative despairJudith Butler, Johns Hopkins University

Steve Goodman

What are the possibilities for multispecies justice? How do social justice struggles intersect with the lives of animals, plants, and other creatures? Leading thinkers in anthropology, geography, philosophy, speculative fiction, poetry, and contemporary art answer these questions from diverse grounded locations. In America, Indigenous peoples and prisoners are decolonizing multispecies relations in unceded territory and carceral landscapes. Small justices are emerging in Tanzanian markets, near banana plantations in the Philippines, and in abandoned buildings of Azerbaijan as people navigate relations with feral dogs, weeds, rats, and pesticides. Conflicts over rights of nature are intensifying in Colombia’s Amazon. Specters of justice are emerging in India, while children in Micronesia memorialize extinct bird species. Engaging with ideas about environmental justice, restorative justice, and other species of justice, The Promise of Multispecies Justice holds open the possibility of flourishing in multispecies worlds, present and to come. Contributors. Karin Bolender, Sophie Chao, M. L. Clark, Radhika Govindrajan, Zsuzsanna Dominika Ihar, Noriko Ishiyama, Eben Kirksey, Elizabeth Lara, Jia Hui Lee, Kristina Lyons, Michael Marder, Alyssa Paredes, Craig Santos Perez, Kim TallBear

L. Ogden, Billy Hall, K. Tanita

Yuk Hui

This book employs recursivity and contingency as two principle concepts to investigate into the relation between nature and technology, machine and organism, system and freedom. It reconstructs a trajectory of thought from an Organic condition of thinking elaborated by Kant, passing by the philosophy of nature (Schelling and Hegel), to the 20th century Organicism (Bertalanffy, Needham, Whitehead, Wiener among others) and Organology (Bergson, Canguilhem, Simodnon, Stiegler), and questions the new condition of philosophizing in the time of algorithmic contingency, ecological and algorithmic catastrophes, which Heidegger calls the end of philosophy. The book centres on the following speculative question: if in the philosophical tradition, the concept of contingency is always related to the laws of nature, then in what way can we understand contingency in related to technical systems? The book situates the concept of recursivity as a break from the Cartesian mechanism and the drive of system construction; it elaborates on the necessity of contingency in such epistemological rupture where nature ends and system emerges. In this development, we see how German idealism is precursor to cybernetics, and the Anthropocene and Noosphere (Teilhard de Chardin) point toward the realization of a gigantic cybernetic system, which lead us back to the question of freedom. It questions the concept of absolute contingency (Meillassoux) and proposes a cosmotechnical pluralism. Engaging with modern and contemporary European philosophy as well as Chinese thought through the mediation of Needham, this book refers to cybernetics, mathematics, artificial intelligence and inhumanism.

Serife Tekin, Carmen Fies, Chris Packham

Capitalizing on the enthusiasm about space science in the general public, our goal as an interdisciplinary group of scholars is to design and teach a new team-taught interdisciplinary course, "Philosophy and Science of Space Exploration (PoSE)" at the University of Texas at San Antonio (UTSA) where we currently teach. We believe that this course will not only help overcome disciplinary silos to advance our understanding of space and critically examine its ethical ramifications, but also will better educate the public on how science works and help overcome the science skepticism that has unfortunately become more prominent in recent years. In what follows, we first juxtapose two seemingly contradictory trends: increased interest in space science on the one hand and increased skepticism about and distrust in science on the other. We then turn to how our anticipated Philosophy and Science of Space Exploration (PoSE) course will develop tools that could dismantle distrust in science while also enhancing the scientific and philosophical understandings of space science. We explain the content and the questions we will examine in POSE and conclude with how we will measure our success and progress.

William Seymour

The home is often the most private space in people's lives, and not one in which they expect to be surveilled. However, today's market for smart home devices has quickly evolved to include products that monitor, automate, and present themselves as human. After documenting some of the more unusual emergent problems with contemporary devices, this body of work seeks to develop a design philosophy for intelligent agents in the smart home that can act as an alternative to the ways that these devices are currently built. This is then applied to the design of privacy empowering technologies, representing the first steps from the devices of the present towards a more respectful future.

Gustavo Caponi

No século XIX, o progresso da biologia evolutiva esteve pautado por dois programas de pesquisa: um de desenvolvimento mais amplo e com maior reconhecimento institucional, que foi o programa filogenético; e outro, o programa adaptacionista, cujo desenvolvimento foi mais restrito ou até mesmo marginal. Fritz Müller (1822-1897) contribuiu para ambas as agendas de pesquisa. Sua contribuição pioneira para o programa filogenético, está em seu livro Für Darwin. Instigado pelo próprio Darwin, também obteve resultados de pesquisa que foram marcos para a consolidação do programa adaptacionista, como por exemplo, seus trabalhos sobre mimetismo.

J. Butler