ODSC East 2023
2023-05-09..2023-05-11
-- online --
who it's for
machine learning engineers, data scientists, nlp engineers, mlops engineers, ai researchers, product managers, data engineers
topics
large language model hallucination mitigation; retrieval augmented generation and retrieval; LoRA fine tuning for LLMs; embeddings for semantic search; multilingual BERT classification; private LLM deployment (PrivateGPT); scaling laws and emergent behaviors; interactive explainable AI; interpretability measurement metrics; testing positive semidefiniteness and eigenvalues; sports analytics expected points added; practical MLOps and data versioning
| title | author | length |
|---|---|---|
| Truth Checker: Generative Large Language Models and Hallucinations | Chandra Khatri | 41m |
Claim: a post‑processing system called Truth Checker ties LLM outputs to an enterprise knowledge base and drops hallucinations to under 1%. It pairs a… | ||
| Using Data Science to Better Evaluate American Football Players | Eric Eager, PhD | 33m |
Shows how play‑by‑play charting plus tenth‑of‑a‑second tracking data produce concrete metrics: builds expected points added and a model that yields PFF wins… | ||
| Reasoning in Natural Language | Dan Roth, PhD | 45m |
Current large LMs show fluent but brittle behavior: they hallucinate facts, fail spatial and scheduling puzzles, and miss simple logical set and counting… | ||
| Semantic Search | Nils Reimers | 25m |
Shows that pre-trained transformer embeddings plus a vector database replace messy language-specific hacks by encoding meaning, with a multilingual embedding… | ||
| Winning The Room: Creating And Delivering An Effective Data-Driven Presentation | Bill Franks | 33m |
Practical checklist for live, nontechnical presentations: plan the script first, use Excel/PowerPoint but show only the few numbers that prove the point… | ||
| Testing Positive Semidefiniteness and Eigenvalue Approximation | David P. Woodruff, PhD | 48m |
Provides optimal, concrete algorithms: for L1 testing the procedure uses Krylov/Chebyshev ideas to get O(eps^{-1/3}) matrix‑vector products (and a matching… | ||
| If We Want AI to be Interpretable, We Need to Measure Interpretability | Jordan Boyd-Graber, PhD | 41m |
Argues that interpretability must be measured, not promised, and introduces two concrete approaches: an intruder interpretability metric for topic models that… | ||
| Leverage Reviews Data for Multi Label Topics Classification | Moran Beladev | 38m |
Fine-tuned multilingual BERT-like models classify 240 tourism topics across reviews and QA, outperforming off-the-shelf zero-shot encoders and needing only… | ||
| Scaling Laws, Emergent Behaviors, and AI Democratization | Irina Rish, PhD | 36m |
Argues that neural scaling follows predictable formulas that extend power laws to capture abrupt emergent jumps in ability, and shows a fitted broken neural… | ||
| Solving MLOps from First Principles | Dean Pleban | 42m |
A five-step, practical MLOps framework: first principles reasoning, define the exact problem, list concrete parameters, Google for existing solutions… | ||
| Interactive Explainable AI | Meg Kurdziolek, PhD | 37m |
Argues that interactive XAI uncovers what models actually use and reduces biases by letting users manipulate inputs and see outputs. Demonstrates… | ||
| Mastering PrivateGPT: Tailoring GenAI for your unique applications | Iván Martínez Toro, Daniel Gallego Vico | 58m |
PrivateGPT is an open‑source GenAI framework with ~52k GitHub stars, ~55 contributors and a ~4k Discord community that ships a usable API, a PrivateGPT Python… | ||
| Applying Responsible AI with Open-Source Tools | David Talby, PhD | 42m |
Claims that open-source testing bridges the gap between papers and production with concrete tools and numbers: the ACL 2020 robustness checklist finds typos or… | ||
| Secrets of Successful AI Projects | Pedro Domingos, PhD | 37m |
Many AI projects fail because teams pick the wrong problem or misunderstand what AI can do, so prioritize narrow, well‑defined tasks with lots of data and low… | ||
| Infuse Generative AI in your apps using | Eve Psalti | 24m |
Demo of the Azure OpenAI Service shows the Private GPT API and Python SDK, configuring PrivateGPT for varied deployments and a bring‑your‑own RAG pipeline that… | ||
| Responsible AI In Practice | Minsoo Thigpen, Dr. Mehrnoosh Sameki | 38m |
Open-source tooling (Fairlearn, InterpretML, ErrorAnalysis) plus a web Responsible AI Dashboard and a new RI mitigations library are usable end-to-end and… | ||
| Text and Code Embeddings | Arvind Neelakantan, PhD | 31m |
Explains that OpenAI trains general-purpose text and code embeddings by contrastive training with very large batch sizes, initializing models from GPT/Codex… | ||
| nothing matches | ||