Colloquia and Seminars

Brownian Bridge Diffusion Models (BBDM) offer an appealing approach to image restoration by constructing a stochastic bridge between a clean image and its degraded observation. Although the bridge schedule plays an important role in reconstruction quality, it is typically chosen heuristically, motivating a principled framework for schedule design.

In this talk, an analytical framework for understanding how BBDM schedules affect the distribution of reconstructed images will be presented. Under a Mixture-of-Gaussians prior, we obtain a closed-form posterior and an ideal minimum mean squared error (MMSE) denoiser. We then compare the posterior law with a tractable surrogate for the BBDM reconstruction law. The surrogate preserves the posterior mean but exhibits a covariance deficit, revealing a coordinate-wise tradeoff between reconstruction fidelity, measured by mean squared error, and distributional accuracy, measured by Wasserstein distance.

Building on this analysis, we derive complementary schedule-design objectives and propose schedules based on analytical bounds that are independent of the prior and degradation. Extensive experiments on controlled MoG settings confirm full alignment between theory and practice, and experiments on the FFHQ dataset across inpainting, deblurring, and super-resolution tasks validate the practical value of our schedule-design criteria.

Rollups improve blockchain scalability by aggregating transactions, yet sequencer centralization introduces risks like ordering manipulation and censorship, which are difficult to verify. While current approaches focus on distributed sequencing or state correctness, they overlook the need for ordering accountability. We introduce DOPPIO, a framework for retrospective verification of fair sequencing. DOPPIO employs a dual-commitment scheme, binding the sequencer to both the ordering metadata and execution batch, and extends the base detector with value-weighted and spatial-concentration variants that, aggregated over consecutive batches, detect sustained localized maximal extractable value (MEV) attacks. We characterize the per-pair limit of observer-based detection: sub-margin delay masking of an isolated transaction pair is indistinguishable from network jitter. Validated on real Ethereum mainnet mempool data, DOPPIO provides accountability with negligible communication overhead and no added execution latency.

Modern large language models increasingly require long contexts for reasoning and multi-document tasks, but attention’s quadratic complexity creates a severe computational bottleneck. We present Block Sparse Flash Attention (BSFA), a drop-in replacement that accelerates long-context inference while preserving model quality. Unlike methods that predict importance before computing scores, BSFA computes exact query-key similarities to select the top-k most important value blocks for each query. By comparing per-block maximum scores against calibrated thresholds, we skip approximately 50% of the computation and memory transfers for pruned blocks. Our training-free approach needs only a single calibration step on a small dataset to set the thresholds for each layer and head. We provide a CUDA kernel implementation that can be used as a drop-in replacement for FlashAttention. On Llama-3.1-8B, BSFA achieves up to 1.13× end-to-end speedup on LongBench with only a 1.1% accuracy drop, and up to 1.24× on Needle-in-a-Haystack retrieval at a 1% accuracy drop. The attention kernel itself accelerates by up to 1.38×. We compare BSFA against five recent sparse attention baselines (SpargeAttention, MInference, FlexPrefill, XAttention, and BLASST), and verify the method on Qwen2.5-7B and on A6000 and H100 GPUs. The implementation is available at https://github.com/Danielohayon/Block-Sparse-Flash-Attention.

The modeling of Water Distribution Networks is a subfield of Water Engineering, with impact of billions of lives worldwide, powering communities, cities, and countries.

To handle such massive loads of data intake, software is a necessary tool for engineers, used to describe, model, analyze and interact with this infrastructure.
The de-facto standard for Water Distribution Systems modeling software is EPANET, an ODE solver library used for its suite of computational tools. The library is written in C++, with many Python programming language wrapper libraries made over it, to enable access to high-level modeling.
Even with accessible libraries available to engineers, such tools remain underutilized due to technical or expertise barriers.

In recent years, Large Language Models changed the landscape of software engineering and accessible abilities. More code than ever is AI-generated, and coding agents are used not only by software engineers, but all vocations are using automated code generation.

In this work, we introduce LLM-EPANET, an agent-based framework that enables natural language interaction with EPANET.
The framework combines retrieval-augmented generation (RAG) and multi-agent orchestration to automatically translate user queries into executable code, run simulations, and return structured results, skipping the needed technical depth to write and debug executable code.

We propose WaterAgentBench, a curated set of 69 benchmark queries to evaluate generated code performance across multiple categories relevant to WDS modeling, in tasks ranging from basic question answering on networks to multiple-command hydraulic analysis queries.

Results show that LLMs can effectively support a wide range of modeling tasks, achieving 56–81% accuracy overall and over 90% for simpler queries, highlighting the potential of LLM-based modeling to support broader accessibility of data-driven tools and advance the democratization of water systems modeling through transparent, interactive AI interfaces.