arXiv subjects

Emmanuel Casseau

Publications and source records attributed to Emmanuel Casseau.

3 recordsLinked to original sources

arXiv2026

Clustering and Token Denoising for Faster and More Robust VLMs

Recent Visual-Language Models (VLMs) have enhanced the capabilities of pre-trained LLMs by adding vision tokens alongside text, with approaches like LLaVA showing impressive results. However, the computational burden of processing up to 576 or 729 visual tokens makes edge deployment challenging. While various token pruning techniques require retraining, some are training-free and thus can easily adapt to architecture changes. We introduce ClustRS, a two-part, training-free algorithm for robust token pruning. Its first component is an attention-weighted, clustering algorithm that selects representative tokens from each semantic cluster. The second component, Residual Shrinkage, is a one-pass denoising step on the selected tokens. These training-free lightweight steps make LLaVA ready for real-world data, improving robustness to a wide range of image-noise types and intensities. Experimental results on the ScienceQA-IMG and MM-VET benchmarks show our method outperforms attention- and diversity-based methods by up to 20\% under extreme noise and token conditions (reducing tokens by 97\%, down to 16 tokens) on LLaVA 1.5 7b and achieves exceptional results on LLaVA-OneVision, where we match baseline performance with fewer than one-third of their tokens under mild noise conditions. Our study demonstrates a simple yet powerful alternative to both score-only and diversity-only pruning rules, paving the way for compute-efficient and noise-resilient VLM deployment.

cs.CV↗

arXiv2025

Intelligent4DSE: Optimizing High-Level Synthesis Design Space Exploration with Graph Neural Networks and Large Language Models

High-Level Synthesis (HLS) Design Space Exploration (DSE) is essential for generating hardware designs that balance performance, power, and area (PPA). To optimize this process, existing works often employs message-passing neural networks (MPNNs) to predict quality of results (QoR). These predictors serve as evaluators in the DSE process, effectively bypassing the time-consuming estimations traditionally required by HLS tools. However, existing models based on MPNNs struggle with over-smoothing and limited expressiveness. Additionally, while meta-heuristic algorithms are widely used in DSE, they typically require extensive domain-specific knowledge to design operators and time-consuming tuning. To address these limitations, we propose ECoGNNs-LLMMHs, a framework that integrates graph neural networks with task-adaptive message passing and large language model-enhanced meta-heuristic algorithms. Compared with state-of-the-art works, ECoGNN exhibits lower prediction error in the post-HLS prediction task, with the error reduced by 57.27\%. For post-implementation prediction tasks, ECoGNN demonstrates the lowest prediction errors, with average reductions of 17.6\% for flip-flop (FF) usage, 33.7\% for critical path (CP) delay, 26.3\% for power consumption, 38.3\% for digital signal processor (DSP) utilization, and 40.8\% for BRAM usage. LLMMH variants can generate superior Pareto fronts compared to meta-heuristic algorithms in terms of average distance from the reference set (ADRS) with average improvements of 87.47\%, respectively. Compared with the SOTA DSE approaches GNN-DSE and IRONMAN-PRO, LLMMH can reduce the ADRS by 68.17\% and 63.07\% respectively.

cs.LG↗

arXiv2023

Algorithms with improved delay for enumerating connected induced subgraphs of a large cardinality

The problem of enumerating all connected induced subgraphs of a given order $k$ from a given graph arises in many practical applications: bioinformatics, information retrieval, processor design,to name a few. The upper bound on the number of connected induced subgraphs of order $k$ is $n\cdot\frac{(eΔ)^{k}}{(Δ-1)k}$, where $Δ$ is the maximum degree in the input graph $G$ and $n$ is the number of vertices in $G$. In this short communication, we first introduce a new neighborhood operator that is the key to design reverse search algorithms for enumerating all connected induced subgraphs of order $k$. Based on the proposed neighborhood operator, three algorithms with delay of $O(k\cdot min\{(n-k),kΔ\}\cdot(k\logΔ+\log{n}))$, $O(k\cdot min\{(n-k),kΔ\}\cdot n)$ and $O(k^2\cdot min\{(n-k),kΔ\}\cdot min\{k,Δ\})$ respectively are proposed. The first two algorithms require exponential space to improve upon the current best delay bound $O(k^2Δ)$\cite{4} for this problem in the case $k>\frac{n\logΔ-\log{n}-Δ+\sqrt{n\log{n}\logΔ}}{\logΔ}$ and $k>\frac{n^2}{n+Δ}$ respectively.

cs.DS↗