Senior ML Engineer Model Compression

About the Team

TheCompression and Parityteam in GMsAutonomous Vehicle (AV) Organizationenables repeatable high-velocitymodeldeploymentsthroughprincipled and automatedmodel compressionunder strict safety guarantees.We partner closely with model developers and deployment and infra engineers to ship numerically robust low-latency models to the car blending rigorous analysis withstate-of-the-artmethods and our own innovations.

About the Role

Over time you will help grow and evolve the Compression and Parity functionthrough the following:

• Developing and iterating on quantizationand compressionstrategiesfor ourAV modelsconsideringmodel numerical properties safety and latency constraints andhardware performance andpartnering on deployment of quantized modelsto NVIDIAbased AV hardwarewith our deployment compiler and kernel teams

• Advancing our numerical sensitivity analysesto recommend safecompressionpolicies per op/layer/blockusing AV-relevant metrics(perception trajectoryetc.) to evaluate compressed modelsandcollaborating withEmbodied AItosupportcompression-aware modeling

• Evolvingsensitivity analysis compression and parity toolinginto a connected automated flow that makes lowprecision deployments repeatable reliable and lowtouchwith an emphasis onrobust execution and maintainability

• Bridging the gapbetweenstate-of-the-artmodel compressionresearch andsafety-constrained deploymentwhilemakingstrong technical contributions incross-functional projectsand educating others on best practices

Your Skills & Abilities (Required Qualifications)

• Bachelorsdegreein Computer Science Electrical Engineering Physics Mathematics Data Science / ML or a closely related quantitative field (or equivalent experience)

• 3 yearsof industry experience focused onmodel optimization and deployment with significant handson work inneural network quantization / model compression / efficient inference or relevant experience

• StrongproficiencyinPyTorchandexperience withgraphlevel representations(e.g.PyTorchFX ONNX) for capture and manipulation

• Background innumerical linear algebra and optimization(conditioning spectral properties Jacobians Hessians) and how they relate to quantization robustness

What Will Give YouACompetitive Edge (Preferred Qualifications)

• Mastersor PhD degreein related quantitative fields

• Deep experience withPTQ and QATcompression frameworks(e.g.PT2EModelOpttorchao) andadvanced quantization algorithms(e.g. GPTQ AWQSmoothQuantQuIPSparseGPT)as well as with building or extendingquantization toolchains

• Handson experience designingnumericsobservability and sensitivity toolingintegrated into training or evaluation pipelines (logging ranges saturation quant noise etc.)

• A track recordof collaborationincludingleading cross-functional initiatives andmentoring others

• Experience withadditionalcompression techniquessuch as structured/unstructuredpruninglowrank decomposition orknowledge distillation

• Experience withperceptionand/or transformerbased models(e.g. multiview encoders BEV backbones detection/segmentation headstrajectoryor planning networks) ideally inAV / ADAS

• General understanding ofkernel performance and optimizationfor reduced precision formats

• Direct experience withspecialized hardware accelerators foredge deploymenton tight latency and memory budgets (automotive SoCs robotics platforms or similar)

• Published research opensource contributions orothernotable intellectually curiousworkin quantization compression or efficient inference

• 3 yearsof industry experience focused onmodel optimization and deployment with significant handson work inneural network quantization / model compression / efficient inference or relevant experience

Compensation:The compensation information is a good faith estimate only. It is based on what a successful applicant might be paidin accordance withapplicable state laws. The compensation may not be representative for positionslocatedoutside of New York Colorado California or Washington.

• The salary range for this role:is $128700 to $ actual base salary a successful candidate will be offered within this range will vary based on factors relevant to the position.

• Bonus Potential:An incentivepayprogram offers payouts based on company performance job level and individual performance.

• Benefits:GM offers a variety of health and wellbeing benefit programs. Benefit options include medical dental vision Health Savings Account Flexible Spending Accounts retirement savings plan sickness and accident benefits life insurance paid vacation & holidays tuitionassistanceprograms employeeassistanceprogram GM vehicle discounts and more.

#GM-AV-1