Oscar Moll, PhD

EMNLP 2024 results are out, we got in three papers, congrats to students :) First paper of many, Vatsal, Pranshu, Suyash, Janit, Kunal, Srija and Adnan, Congrats 🎉 🎊 🍾 🎈 Evaluating Concurrent Robustness of Language Models Across Diverse Challenge Sets (main) - Vatsal Gupta, PRANSHU PANDYA, Tushar Kataria, Vivek Gupta, Dan Roth tldr: make models robust on multiple adversarial challenge set simultaneously  arxiv: https://lnkd.in/ei-65Y6g Knowledge-Aware Reasoning over Multimodal Semi-structured Tables (finding) Suyash Vardhan Mathur, Jainit Bafna, Kunal Kartik, Harshita Khandelwal, Manish Shrivastava, Vivek Gupta, Mohit Bansal, Dan Roth tldr: introduce the novel task of reasoning on multimodal tabular data, text + images in a structure (first of a kind) arxiv: https://lnkd.in/e5ycusfn   Unraveling the Truth: Do LLMs really Understand Charts? A Deep Dive into Consistency and Robustness (finding) Srija Mukhopadhyay, Adnan Qidwai, Aparna Garimella, Pritika Ramu, Vivek Gupta, Dan Roth           tldr: VLM although seem very accurate on chart, are not all robust and consistent in performance arxiv: https://lnkd.in/eFFA-sVX We also got an workshop paper, at BlockBoxNLP 2024 @EMNLP, here the details Enhancing Question Answering on Charts Through Effective Pre-training Tasks Ashim Gupta, Vivek Gupta, Shuo Zhang, Yujie He, Ning Zhang , Shalin Shah, Ph.D. tldr: models is not robust & consistent on charts, can we use pre-training on simple auxiliary tasks to enhance model robustness arxiv: https://lnkd.in/egbRHm-4

386

12 Comments

Our #icra2024 paper is now available from #ieee dl. We proposed a new Inverse Reinforcement Learning based method for hand trajectory prediction. Mukund Mitra Gyanig K. Partha Pratim Chakrabarti

18

2 Comments

The parameters of compiler optimizations interact in complex and non-trivial ways, and finding optimal combinations remains an open research challenge. There are several approaches to discovering effective combinations of these optimization parameters. Recently, Michael Canesche (Cadence Design Systems) and Gaurav Verma (Stony Brook University) have had promising results using auto-tuning with coordinate descent. To visualize this process, imagine that each optimization represents a 'dimension' in a search space. The combination of all optimizations and their respective parameters forms a vector in this space. Now, consider that each point in this space is mapped to a real number representing the performance of a kernel (assuming a fixed input). Coordinate descent is a search technique that assumes the performance curve is convex. Based on this assumption, it aims to find the optimal point by varying one optimization parameter at a time. Explore as a Storm, Exploit as a Raindrop: On the Benefit of Fine-Tuning Kernel Schedulers with Coordinate Descent: https://lnkd.in/dgxxfdYR The Droplet Search Algorithm for Kernel Scheduling: https://lnkd.in/dBfQid9z

77

I remember when this all started. Faris - invited me to catch up at our favorite spot around Blok M. I had no idea he’d just wrapped up a procedure at Pondok Indah Hospital! (He really knows how to add drama to our catch-ups.) We’d been on separate paths and projects since "the winding down project", but there he was, reminding me of our dream to ride this AI wave together again. It was a plan we’d hatched with Sahal Zain & Gurnoor Dhillon when the world of AI was still mostly theoretical and jargon-heavy. Somehow, the conversation continued with Gurnoor Dhillon while he was waiting for his meeting with someone who would eventually become one of CalvinBall’s key early adopters. It felt like everything was coming together, like a series of seemingly random dots connecting to form a bigger picture. Seeing how far we’ve come as a team at CalvinBall Technologies is truly exciting. What started as theoretical discussions has now evolved into practical Gen AI solutions that are solving everyday problems and hold endless possibilities to shape the industry. 👨‍🚀 👩‍🚀 Here’s to pushing boundaries and achieving new milestones together! Gurnoor Dhillon, Sandeep Ramesh, Faris -, Sahal Zain, Eva Zuliya, Ratu Nadiah Khairunnisa 🚀 #GenAI #AIInnovation #TechJourney #ArtificialIntelligence #TechTransformation #CalvinballTech #Calvinballers

47

2 Comments

IndicGenBench Google Research India recently released IndicGenBench, a multilingual benchmark to evaluate generation capabilities of LLMs on 29 Indic languages spanning 13 writing scripts and 4 language families. Extended the datasets in Cross-lingual Summarization, Machine Translation, Multi-lingual Question Answering, and Cross-lingual Question Answering, the team has collected human translations for English examples into target Indic languages, thereby extending the scope and applicability of evaluation metrics in this domain. One of the key insights from their study is the analysis of token fertility across all Indic languages within IndicGenBench. Token fertility, representing the average number of sub-words that a word is broken down into by the tokenizer, varies significantly across languages Some languages have simple breakdowns, while others are more complex. Now, why does this matter? Well, it affects how well the language models work. Languages with more complex breakdowns might struggle because they can't use as many examples to learn from. They found that languages with simpler breakdowns can use more examples effectively compared to those with complex breakdowns. #LLMs #GenAI #IndicGenBench #AI #IndicLanguages #IndicDatasets #multilingual

15

1 Comment

Released a new paper on how to use LLMs to generate unit tests for PLC code. Getting valid test inputs worked well in a few examples but the assertions were often hallucinated. Needs more research, but already can save developers testing efforts. Virendra Ashiwal, Soumyadip Bandyopadhyay, Chandrika K R, and me from ABB conducted prompt engineering and created a test generator and test execution framework for IEC 61131-3 ST code based on open-source (MATIEC, OpenPLC, GCC). Used several open-source function blocks (OSCAT) for experimentation. Our paper was just accepted at IEEE ETFA 2024: https://2024.ieee-etfa.org 8-page report on arXiv: https://lnkd.in/dn3dZguQ Github repo with the code: https://lnkd.in/dxb7JhBC

167

16 Comments

I thought it was very interesting to hear Andrej Karpathy talk at GPU MODE IRL 2024 about llm.c and the conversion of PyTorch code into C/CUDA, as well as the potential implications of this work for future LLM progress. The idea of converting high-level deep learning frameworks like PyTorch into low-level, highly efficient C/CUDA code could drastically improve model performance and execution speed, especially in specialized environments. It will be exciting to see how software engineering (SWE) practices evolve over the coming years as this becomes more widely adopted. One of the most intriguing aspects is the potential for a model to generate a specific kernel for a use case at someone's company in code that has no dependencies and could be fully optimized for that specific task. This capability could not only make models more efficient by reducing overhead but also streamline the workflow for developers, allowing them to focus on higher-level design and problem-solving. Essentially, it introduces a whole new paradigm for how we think about integrating AI into SWE roles. Instead of relying on pre-built libraries or packages, engineers might one day leverage AI models to dynamically create specialized, hyper-optimized code that meets their unique needs, reducing both the time and complexity of development. This shift could also lead to a greater level of automation, where AI models help developers by taking over tasks like performance optimization or low-level code generation, potentially reshaping the way we approach software development as a whole. You can listen to the keynote here: https://lnkd.in/eqJYS8Gb #AI #MachineLearning #DeepLearning #PyTorch #CUDA #SoftwareEngineering #C #LLM

2

At EMNLP, Amazon researchers showed how to “detoxify” an LLM's output by using an auxiliary model to estimate the distribution closest to the LLM's current distribution that satisfies toxicity constraints and penalizing the gap between the two. #EMNLP2024 #LLM #LargeLanguageModels

30

A recent breakthrough titled "Matrix Multiplication-Free LLMs" demonstrates a huge advancement in the area of Large Language Models (LLMs) by reducing computational costs. The authors have eliminated MatMul operations from LLMs, claiming to 10 times reduction in memory usage and a 25.6% increase in training speed, all while maintaining strong performance at billion-parameter scales. Paper link: https://lnkd.in/ggph8qXc #AI #machinelearning #deeplearning #LLMs

15

Brain Teaser #11: What mathematical arguments can explain how **Layer Normalization** enhances training stability in large language models (LLMs) and Transformer architectures, particularly in mitigating issues like vanishing and exploding gradients? Answer: Training destabilization in neural networks, particularly due to gradient explosion or vanishing, typically occurs because of the high condition numbers of weight matrices during backpropagation. These large condition numbers lead to unstable gradient scaling, either amplifying gradients excessively or diminishing them, hindering effective learning. How Layer Normalization overcomes and brings stability in training is briefly discussed here: https://lnkd.in/gmWm42tQ #MachineLearning #DeepLearning #Transformers #LargeLanguageModels #LayerNormalization #GradientStability #AIResearch #DataScience #NeuralNetworks #AIInsights

16

5 Comments

🔥 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈 𝐚𝐧𝐝 𝐋𝐋𝐌 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 - 𝐓𝐨𝐩 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧𝐬 🔥 🤖 I've conducted a mock interview with my student, Bheesetti Anand, focused on Generative AI and LLM (Large Language Models) interview questions. This session is designed to benefit other students preparing for similar interviews, helping them secure their dream roles in the field of Generative AI. 📔 Here’s why watching the video can help you make a 𝐬𝐮𝐜𝐜𝐞𝐬𝐬𝐟𝐮𝐥 𝐜𝐚𝐫𝐞𝐞𝐫 𝐬𝐰𝐢𝐭𝐜𝐡 𝐭𝐨 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈: ✅ 𝐂𝐨𝐦𝐩𝐫𝐞𝐡𝐞𝐧𝐬𝐢𝐯𝐞 𝐂𝐨𝐯𝐞𝐫𝐚𝐠𝐞: Detailed exploration of top interview questions related to Generative AI and Large Language Models. ✅ 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐚𝐥 𝐈𝐧𝐬𝐢𝐠𝐡𝐭𝐬: Real-world examples and practical applications to understand complex concepts better. ✅ 𝐄𝐱𝐩𝐞𝐫𝐭 𝐆𝐮𝐢𝐝𝐚𝐧𝐜𝐞: Tips and strategies to approach and answer interview questions effectively. ✅ 𝐂𝐚𝐫𝐞𝐞𝐫 𝐏𝐫𝐞𝐩𝐚𝐫𝐚𝐭𝐢𝐨𝐧: Essential knowledge and skills to prepare for interviews and stand out to potential employers. 📽 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 𝐥𝐢𝐧𝐤: https://lnkd.in/g_nkP7fs #generativeai #llm #interview #jobpreparation #mockinterview

14

2 Comments

These look like a great set of courses. Looking at the curriculum if you want a deeper understanding of transformer-llms and multimodal systems you should explore them. Plus its Carnegie Mellon University so you can't go wrong. I believe you need to be C++, CUDA and Python proficient.

10

1 Comment

OpenCoder: An Open-Source Code LLM Redefining Model Efficiency 🌟 The OpenCoder team has launched a game-changer in the world of code language models. Unlike many LLMs that focus on sheer size, OpenCoder brings something unique to the table: a data-focused training approach. By putting quality before volume, OpenCoder’s 1.5B and 8B models rival some of the biggest players, like DeepSeekCoder and QwenCoder, with only a fraction of their token count. The core strength of OpenCoder lies in its meticulous data processing pipeline and two-stage fine-tuning. With a training dataset composed of 2.5 trillion tokens, 90% raw code and 10% code-related web data, OpenCoder refines every aspect of its corpus. Advanced filtering and deduplication strategies, applied to 607 programming languages, optimize model efficiency and precision, ensuring high-quality data drives its training. Key Features: - Complete Training Transparency: Open-source datasets and training pipelines enable reproducibility and further research. - Instruction Fine-Tuning: Two-stage tuning sharpens both core CS concepts and real-world coding skills, prepping OpenCoder for production-level code generation. - Available on Ollama: Accessibility in deployment means more avenues for experimentation and improvement. OpenCoder's data-driven strategy showcases the potential of a leaner approach. By focusing on quality data rather than quantity alone, OpenCoder achieves file-level deduplication that outperforms traditional repo-level methods, making it both efficient and robust. 🚀 Read more here: https://lnkd.in/eEnG3yJA #OpenSource #MachineLearning #CodeLLM #AI #ModelEfficiency #DataQuality #Innovation #CodeGeneration #OpenCoder #TechNews #AIResearch #Programming #LargeLanguageModels #ArtificialIntelligence

1

Excited to share our DistiLLM accepted at #ICML2024. DistiLLM is a better (performance), faster (training efficiency), and stronger (theorem) distillation method to significantly enhance the performance of SLMs. #efficientAI #slm

26

Shashank Sonkar and colleagues are back with another great paper and matching GitHub repo, "Pedagogical Alignment of Large Language Models". You can read the paper on arXiv here - https://lnkd.in/enZfCJgi, and browse the code here - https://lnkd.in/gyW5hSzp. They demonstrate an interesting approach to generating synthetic data to train open weights models to behave more pedagogically, as well as developing some new evaluation metrics. I think there's lots here for those of us interested in this kind of work to learn from and build on. (If synthetic data generation is new do you, check out this post from answer.ai earlier this week - https://lnkd.in/ecsG7R4b.)

38

4 Comments

The paper "Scalable MatMul-free Language Modeling" by Rui-Jie Zhu et al. introduces a method to eliminate matrix multiplication (MatMul) operations in large language models (LLMs), significantly reducing memory usage and computational cost. Their MatMul-free models perform comparably to state-of-the-art transformers, scaling effectively up to 2.7 billion parameters. They achieved memory reductions of up to 61% during training and over 10x during inference by using a custom FPGA hardware solution, which highlights the efficiency of lightweight operations for future language model accelerators. For further details, you can read the full paper https://lnkd.in/eZidqJuk

2

The paper (2020) by Jared Williard et al on ‘Integrating Physics-based modeling with machine learning’ makes for an insightful read. ⬇️ What they provide through this paper are the objectives of Physics-#ML integration, which clearly are: 👉 finding computationally inexpensive representations of more complex #models 👉 discovering governing equations 👉 solving PDEs 👉 uncertainty quantification 👉 #data generation Physics-based approaches for #data generation often rely on running physical simulations or experiments, which tend to be very time consuming. Also, they are restricted by what can be produced by only physics-based #models. Hence there is an increasing interest in deep learning #DL approaches. Neural network approaches like #GAN, #VAE learn #data distributions in unsupervised settings and have the potential to generate new #data beyond what could be produced by traditional approaches. The generative adversarial network #GAN is one #DL architecture that extracts representative latent embeddings from observed #data distributions; #DL methods are known to approximate distributions very well. The #data generated by capturing latent embeddings are similar to observed #data. The adversarial component of #GAN consists of a 2-part framework namely, generative network and discriminative network; the generator’s objective is to generate sample #data and the discriminator determines if the sample came from training/observed #data or the generator. The generative and discriminative networks thus compete against each other and a #GAN learns to generate new samples.

5

1 Comment

Great to see the newly announced Bamba-9B, an inference-efficient Hybrid Mamba2 model 🐍 trained by IBM, Princeton, CMU, and UIUC on completely open data used PyTorch FSDP to train these novel architecture models and they are integrating inference with vLLM, which recently joined the PyTorch Ecosystem.

167