UC Berkeley engineer working on autonomy, controls, and machine learning.https://parham-sharaf.github.io/en-ushttps://parham-sharaf.github.io/projects/personal-website/https://parham-sharaf.github.io/projects/personal-website/Initial repository for a personal website. Currently a minimal public scaffold with the GitHub repo in place and ready for iterative development.Tue, 21 Apr 2026 00:00:00 GMTWeb Developmentportfoliopersonal-websitewebhttps://parham-sharaf.github.io/projects/modern-nlp/https://parham-sharaf.github.io/projects/modern-nlp/Four paradigm shifts in one semester: IBM Model 1 → attention-based NMT → transformer parsing → LLM fine-tuning → CLIP multimodal retrieval with pragmatic reasoning. Each technique subsumes and extends the previous.Tue, 10 Dec 2024 00:00:00 GMTNatural Language Processingnlpmachine-translationattentiontransformersllmsmultimodalpragmaticshttps://parham-sharaf.github.io/projects/neural-radiance-fields/https://parham-sharaf.github.io/projects/neural-radiance-fields/Training an MLP to represent a 3D scene as a continuous function from (x, y, z, θ, φ) to (RGB, density). Volume rendering turns the field back into images; the field itself is the 3D model.Tue, 10 Dec 2024 00:00:00 GMTComputer Visionnerfneural-renderingvolumetric-rendering3d-reconstructionhttps://parham-sharaf.github.io/projects/vit-masked-autoencoder/https://parham-sharaf.github.io/projects/vit-masked-autoencoder/ViT classifier achieving 73.5% on CIFAR-10, then self-supervised MAE pretraining boosts finetuned accuracy to 76.8%. Full implementation of patchify, attention pooling, and mask reconstruction.Mon, 25 Nov 2024 00:00:00 GMTDeep Learningvision-transformermasked-autoencoderself-supervisedtransformershttps://parham-sharaf.github.io/projects/diffusion-models/https://parham-sharaf.github.io/projects/diffusion-models/Sampling from DeepFloyd IF — CFG, SDEdit, inpainting, visual anagrams. Then training a time- and class-conditioned U-Net from scratch on MNIST to learn the diffusion process end-to-end.Wed, 20 Nov 2024 00:00:00 GMTComputer Visiondiffusiongenerative-modelsstable-diffusionunethttps://parham-sharaf.github.io/projects/transformer-summarization/https://parham-sharaf.github.io/projects/transformer-summarization/Self-attention, multi-head attention, and encoder-decoder architecture implemented from scratch. Trained on CNN/DailyMail achieving 35.1 ROUGE-L, outperforming LSTM baseline by 60%.Fri, 15 Nov 2024 00:00:00 GMTDeep Learningtransformersattentionsummarizationnlphttps://parham-sharaf.github.io/projects/rnn-sequence-modeling/https://parham-sharaf.github.io/projects/rnn-sequence-modeling/Recurrent networks from scratch — forward pass, backpropagation through time, and gradient flow analysis. Vectorized NumPy implementation validated to 5e-5 tolerance.Sat, 02 Nov 2024 00:00:00 GMTDeep Learningrnnlstmsequence-modelingbackpropagationhttps://parham-sharaf.github.io/projects/auto-panorama/https://parham-sharaf.github.io/projects/auto-panorama/Building a panorama pipeline from scratch — Harris corner detection, Adaptive Non-Maximal Suppression, feature matching, RANSAC for homography estimation, and Laplacian-pyramid blending.Mon, 28 Oct 2024 00:00:00 GMTComputer Visionpanoramafeature-detectionransachomographyhttps://parham-sharaf.github.io/projects/deep-learning-scratch/https://parham-sharaf.github.io/projects/deep-learning-scratch/Backpropagation, BatchNorm, Dropout, and CNNs implemented from first principles in NumPy — then PyTorch deployment achieving 74.8% on CIFAR-10.Tue, 15 Oct 2024 00:00:00 GMTDeep Learningneural-networksbackpropagationconvolutional-networksbatch-normalizationhttps://parham-sharaf.github.io/projects/face-morphing/https://parham-sharaf.github.io/projects/face-morphing/Smooth warping between two faces via point correspondences, Delaunay triangulation, and affine warps per triangle. Plus: population mean faces and caricature generation by extrapolation.Sat, 12 Oct 2024 00:00:00 GMTComputer Visionface-morphingdelaunay-triangulationaffine-warpingimage-interpolationhttps://parham-sharaf.github.io/projects/filters-frequencies/https://parham-sharaf.github.io/projects/filters-frequencies/Working in the frequency domain to extract edges, create hybrid images that change meaning with viewing distance, and blend images seamlessly via Laplacian pyramids. Ends with the famous 'oraple.'Wed, 25 Sep 2024 00:00:00 GMTComputer Visionimage-filtersfrequency-domainhybrid-imagespyramid-blendinghttps://parham-sharaf.github.io/projects/prokudin-gorskii/https://parham-sharaf.github.io/projects/prokudin-gorskii/Reconstructing color from Sergei Prokudin-Gorskii's glass plate negatives (captured 1907–1915) using image pyramids and normalized cross-correlation alignment.Tue, 10 Sep 2024 00:00:00 GMTComputer Visionimage-alignmentcomputational-photographyhistorical-photographyhttps://parham-sharaf.github.io/projects/pacman-ai/https://parham-sharaf.github.io/projects/pacman-ai/End-to-end AI agent built across four paradigms: A* pathfinding, Minimax/Expectimax game trees, HMM + particle filter tracking, and Q-learning. Each piece solves one Pacman problem; together they cover the full classical AI curriculum.Mon, 05 Aug 2024 00:00:00 GMTArtificial Intelligencesearchminimaxhmmparticle-filterq-learningmdphttps://parham-sharaf.github.io/projects/image-geolocation/https://parham-sharaf.github.io/projects/image-geolocation/Computer vision system predicting photo locations from visual features. Combines k-nearest neighbors with regression models, achieving 127km median error on global street-view dataset.Wed, 01 May 2024 00:00:00 GMTMachine Learningcomputer-visiongeolocationk-nearest-neighborsimage-processinghttps://parham-sharaf.github.io/projects/parallel-convolve/https://parham-sharaf.github.io/projects/parallel-convolve/Hand-optimized SIMD kernels with parallel tiling achieve 17× speedup over naive implementation. Deep dive into vectorization, memory patterns, and performance engineering.Wed, 01 May 2024 00:00:00 GMTParallel Computingsystemsperformancesimdhttps://parham-sharaf.github.io/projects/cs61cpu/https://parham-sharaf.github.io/projects/cs61cpu/A complete 2-stage pipelined RV32I processor built from first principles in Logisim — ALU, register file, control unit, and memory system, all hand-wired from basic gates.Mon, 15 Apr 2024 00:00:00 GMTComputer Architecturearchitecturehardwareriscvhttps://parham-sharaf.github.io/projects/movie-recommender/https://parham-sharaf.github.io/projects/movie-recommender/SVD-based matrix factorization system with 0.89 RMSE on MovieLens dataset. Handles sparse ratings, cold start problems, and scales to 100k+ users through optimized gradient descent.Fri, 05 Apr 2024 00:00:00 GMTMachine Learningrecommender-systemsmatrix-factorizationsvdoptimizationhttps://parham-sharaf.github.io/projects/decision-trees/https://parham-sharaf.github.io/projects/decision-trees/From-scratch implementation of decision trees with pruning, random forests, and AdaBoost. Comprehensive analysis of overfitting, feature selection, and ensemble performance on real datasets.Sun, 10 Mar 2024 00:00:00 GMTMachine Learningdecision-treesensemble-methodsrandom-forestadaboosthttps://parham-sharaf.github.io/projects/riscv-classifier/https://parham-sharaf.github.io/projects/riscv-classifier/A complete 2-layer MLP for digit classification — matrix multiply, ReLU, argmax, file I/O, and all infrastructure — written entirely in hand-coded RISC-V assembly without any library calls.Fri, 01 Mar 2024 00:00:00 GMTSystems Programmingsystemsmlassemblylow-levelhttps://parham-sharaf.github.io/projects/neural-network-scratch/https://parham-sharaf.github.io/projects/neural-network-scratch/Pure NumPy implementation achieving 99.6% MNIST accuracy through optimized gradient descent, backpropagation, and regularization techniques.Thu, 15 Feb 2024 00:00:00 GMTMachine Learningneural-networksgradient-descentbackpropagationhttps://parham-sharaf.github.io/projects/entropy-wordle-solver/https://parham-sharaf.github.io/projects/entropy-wordle-solver/Information-theoretic greedy solver that picks each guess to maximize expected entropy over the remaining word set — averaging 3.92 guesses across 300+ games.Wed, 15 Nov 2023 00:00:00 GMTMachine Learningentropyoptimizationinformation-theoryhttps://parham-sharaf.github.io/projects/mcmc-cipher-decoder/https://parham-sharaf.github.io/projects/mcmc-cipher-decoder/Metropolis-Hastings breaking substitution ciphers against bigram frequencies — watch garbled text slowly resolve into English over MCMC iterations.Sat, 28 Oct 2023 00:00:00 GMTMachine Learningmcmccryptographyprobabilityhttps://parham-sharaf.github.io/projects/byow/https://parham-sharaf.github.io/projects/byow/A sophisticated 2D world generator with dynamic lighting, line-of-sight exploration, custom tilesets, and interactive minimaps — infinite deterministic worlds from a single seed.Mon, 01 May 2023 00:00:00 GMTGame Developmentgamesproceduralgraphicsui-designhttps://parham-sharaf.github.io/projects/ngordnet/https://parham-sharaf.github.io/projects/ngordnet/Interactive tool fusing WordNet's semantic graph with Google Books NGram data to explore how language evolves — query 'transportation' and watch centuries of linguistic change unfold.Sat, 15 Apr 2023 00:00:00 GMTNLP & Data Engineeringnlpdata-vizgraphslinguisticshttps://parham-sharaf.github.io/projects/pocket-planet/https://parham-sharaf.github.io/projects/pocket-planet/A 100×100 world carved from Perlin noise, then colonized by simulated plants that mutate, compete, and converge on the terrain they're fittest for.Sat, 01 Apr 2023 00:00:00 GMTProcedural Generationproceduralsimulationevolution