Aman Chadha @i_amanchadha Twitter profile

Pinned Tweet

Aman Chadha

10 months

Thanks @Analyticsindiam for featuring me and our research on Indic LLMs, in collaboration with the amazing team at IIT Patna!

Meet the AI Expert Building Indic LLMs with IITs

AIM got in touch with the researcher from Stanford University, Aman Chadha, who is currently working on building a Indic medical LLM.

analyticsindiamag.com

2

4

12

Last Seen Profiles

@Stephannie49360

@musayoshi163611

@JURINA141597853

@xieborubok8441

@DJamarl20735

@stw46

@oimosan136510

@yukisakiyo7999

@machidamit1891

@C2y31qmnCWLobuU

@stw46

@papmantan69_

@galery_basah10

@hankoyama

@6y5m129689

@nayuwdogun67834

@excell305

@hankoyama

@VienZubin2431

@SmeecheehyTCj

@seikansan

@stw_pdg

@DumbKami_2

@papmantan69_

@Adrian_tzz1

@papmantan69_

@shinaritij83049

@hankoyama

@waoASA

@ppichchi11888

@dorimukapu85657

@KirthetCM6

@maylanhowat

@sw320358595

@MikaJie15846

@etoyan199513

Aman Chadha

@i_amanchadha

2 years

📚 Natural Language Processing from Stanford University: Distilled Notes 👉🏼 - NLP is one of the most popular #AI domains, widely used from language translation to auto-complete to voice assistants. - Presenting notes from Stanf…

Aman Chadha on LinkedIn: #ai #artificialintelligence #machinelearning #ml #transformers… | 18...

📚 Natural Language Processing from Stanford University: Distilled Notes 👉🏼 http://nlp.aman.ai - NLP is one of the most popular #AI domains, widely used… | 18 comments on LinkedIn

www.linkedin.com

2

69

275

Aman Chadha

@i_amanchadha

5 months

📚 Ilya Sutskever's Top 30 AI Papers • - Ilya Sutskever shared a list of 30 papers with John Carmack, saying, “If you really learn all of these, you’ll know 90% of what matters today.” - In this article on , we have reviewed all of

1

8

45

Aman Chadha

@i_amanchadha

4 months

🚀 Launching - your one-stop shop for free AI resources: - This project has been a work in progress for over a year, and I'm thrilled to finally share it with you. - is a comprehensive platform offering a

0

2

21

Aman Chadha

@i_amanchadha

11 months

🎉 Thrilled to announce that we received an Outstanding Paper Award at #EMNLP2023 ! 🔷Counter Turing Test (CT^2): AI-Generated Text Detection is Not as Easy as You May Think - Introducing AI Detectability Index ➡️ @emnlpmeeting #ArtificialIntelligence #AI

3

2

22

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper surveying Multimodal AI Architectures -- with a comprehensive taxonomy and analysis of their pros/cons & applications in any-to-any modality model development ➡️ 𝐂𝐨𝐦𝐩𝐫𝐞𝐡𝐞𝐧𝐬𝐢𝐯𝐞 𝐓𝐚𝐱𝐨𝐧𝐨𝐦𝐲: First work to explicitly identify and categorize

0

5

20

Aman Chadha

@i_amanchadha

4 months

I’m #hiring for full-time Generative AI roles (Applied Scientists and Deep Learning Architects) in Amazon Web Services ( @awscloud ). We're seeking exceptional talent skilled in deep learning, eager to work on cutting-edge AI, and interested in building the next generation of

5

3

20

Aman Chadha

@i_amanchadha

5 months

🧠 All you need to know about Vision-Language Models (VLMs) / Multimodal Large Language Models (MLLMs) • • Overview • Architecture • Architecture of Vision-Language Models • Examples of Popular VLMs and Their Architectural Choices • Differences from

0

3

19

Aman Chadha

@i_amanchadha

7 months

🗄️Retrieval Augmented Generation (RAG) in LLMs • 🔹RAG Pipeline 🔹Benefits of RAG 🔹RAG vs Fine-tuning 🔹Vector DB Feature Matrix 🔹Component-Wise Evaluation 🔹Multimodal RAG 🔹Related Papers: HyDE, FLARE, Self/Corrective/Multihop-RAG, etc. #GenAI #LLMs

0

5

20

Aman Chadha

@i_amanchadha

8 months

🧠 Primer on Word Embeddings • 🔹 Motivation 🔹 Distributional Semantics 🔹 Conceptual Framework 🔹 Word Embedding Techniques (Count-based, Co-occurence-based, Contextualized) ➡️ TF-IDF, BM25, Word2Vec, GloVe, fastText, BERT #AI #MachineLearning #LLMs

1

4

16

Aman Chadha

@i_amanchadha

7 months

📖Everything NLP Primers • ➡️Curated NLP Primers that focus on topics such as: 🔹Embeddings 🔹LLMs 🔹VLMs 🔹RAG 🔹PEFT 🔹RLHF/Alignment 🔹Context Extension 🔹Eval Metrics 🔹Tokenization 🔹Token Sampling 🔹BERT 🔹GPT 🔹Transformers 🔹MoE 🔹SSMs #GenAI

0

5

18

Aman Chadha

@i_amanchadha

5 months

I’m #hiring for full-time Generative AI roles in @awscloud (Amazon Web Services). We're seeking exceptional talent skilled in deep learning, eager to work on cutting-edge AI, and interested in building the next generation of AI-powered solutions (powered by LLMs, VLMs, Diffusion

4

0

18

Aman Chadha

@i_amanchadha

5 months

I’m #hiring for full-time Generative AI roles in @awscloud (Amazon Web Services). We're seeking exceptional talent skilled in deep learning, eager to work on cutting-edge AI, and interested in building the next generation of AI-powered solutions (powered by LLMs, VLMs, Diffusion

0

3

18

Aman Chadha

@i_amanchadha

5 months

🤖 LLM Alignment Primer (𝐑𝐋𝐇𝐅, 𝐑𝐋𝐀𝐈𝐅, 𝐃𝐏𝐎, 𝐊𝐓𝐎, 𝐆𝐏𝐎, 𝐂𝐏𝐎, 𝐈𝐏𝐎, 𝐈𝐂𝐃𝐏𝐎, 𝐎𝐑𝐏𝐎, 𝐬𝐃𝐏𝐎, 𝐑𝐒-𝐃𝐏𝐎, 𝐒𝐢𝐦𝐏𝐎, 𝐃𝐢𝐟𝐟𝐮𝐬𝐢𝐨𝐧-𝐃𝐏𝐎) 🔗 - Reinforcement Learning with Human Feedback (RLHF) is used to align LLM behavior

0

7

15

Aman Chadha

@i_amanchadha

9 months

📝Announcing our survey paper covering 30+ Prompt Engineering techniques 🔹"A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications" 🔹In collaboration with IIT Patna ➡️ @IITPAT #ArtificialIntelligence #LLM #ChatGPT

2

14

Aman Chadha

@i_amanchadha

4 months

🎨 Diffusion Models Primer • - Diffusion models are all the rage right now owing to their extraordinary applications in text-to-image applications, which has given AI wings to generate art from text prompts. - Here’s my primer on diffusion models that

1

5

14

Aman Chadha

@i_amanchadha

5 months

📈 Mixture-of-Experts (MoE) Primer • - MoE enhances model performance by dynamically selecting specialized subnetworks for different inputs, improving efficiency and scalability. This approach reduces computational cost while maintaining high accuracy,

0

2

14

Aman Chadha

@i_amanchadha

7 months

📖Everything NLP Primers • ➡️Curated NLP Primers that focus on important topics such as: 🔹Word Embeddings 🔹LLMs 🔹VLMs 🔹RAG 🔹PEFT 🔹RLHF 🔹Eval Metrics 🔹Tokenization 🔹Token Sampling 🔹BERT 🔹GPT 🔹Transformers #ArtificialIntelligence #GenAI

0

14

Aman Chadha

@i_amanchadha

9 months

I'm #hiring for full-time Generative AI roles in Amazon AWS. We're seeking exceptional talent interested in developing the next generation of AI-powered solutions. If this is you, get in touch! #ArtificialInteligence #MachineLearning #AI #Jobs

1

0

14

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper on Energy-Based World Models (EBWM), an architecture that emulates facets of human cognition to improve the scalability and performance of autoregressive world models across Computer Vision and Natural Language Processing. ➡️ 𝐍𝐨𝐯𝐞𝐥

1

2

13

Aman Chadha

@i_amanchadha

7 months

📝 All you need to know about LLM Prompt Engineering • ⚡️ This primer presents 30+ flavors of Prompt Engineering along with their use-cases. #ArtificialIntelligence #GenAI #LLMs

0

1

13

Aman Chadha

@i_amanchadha

8 months

📝 Announcing our new paper that proposes a framework to enhance causal reasoning & explainability in LLMs 🔹"Cause and Effect: Can Large Language Models Truly Understand Causality?" 🔹With @CarnegieMellon , @UNTsocial , @rpi , and @UMassAmherst 🔗 #AI #LLM

0

1

14

Aman Chadha

@i_amanchadha

10 months

🗄️ Retrieval Augmented Generation (RAG) in LLMs • 🔹 RAG Pipeline 🔹 Benefits of RAG 🔹 RAG vs. Fine-tuning 🔹 Vector DB Feature Matrix 🔹 Component-Wise Evaluation 🔹 Multimodal RAG 🔹 Related Papers #LLMs #ArtificialIntelligence #MachineLearning #AI

3

0

12

Aman Chadha

@i_amanchadha

4 months

📈 Mixture-of-Experts (MoE) Primer • - MoE enhances model performance by dynamically selecting specialized subnetworks for different inputs, improving efficiency and scalability. This approach reduces computational cost while maintaining high accuracy,

1

3

12

Aman Chadha

@i_amanchadha

4 months

📖 The A-to-Z of Large Language Models (LLMs) • - What are Embeddings? - Contextualized vs. Non-Contextualized Embeddings - How do LLMs work? - LLM Training Steps - Computing Similarity between Embeddings (Dot Product Similarity, Geometric Intuition,

0

3

12

Aman Chadha

@i_amanchadha

3 months

📚 All the Math fundamentals for AI (including Backpropagation primers) - Don’t let the math behind AI concepts hinder you from understanding what goes on under-the-hood as you train your neural network. Think in terms of first principles to develop an intuition of what’s going

0

1

13

Aman Chadha

@i_amanchadha

6 months

📝 Looking to learn about Hallucination in LLMs? ✅ Our tutorial at @LrecColing 2024 will offer an introduction to the issue of hallucination, present a taxonomy of categories, and cover hallucination detection and mitigation methods. 🔗 Tutorial Schedule:

1

3

11

Aman Chadha

@i_amanchadha

7 months

🏅Recommender Systems' Primers • 🔹Candidate Generation & Retrieval 🔹Candidate Ranking & Re-ranking 🔹Cold Start Problem 🔹Position Bias 🔹Content Moderation 🔹Evaluation Metrics and Loss 🔹Multi-Armed Bandits 🔹GNNs for RecSys 🔹Transformers for RecSys

1

2

11

Aman Chadha

@i_amanchadha

6 months

🎉 Happy to share that the Wikipedia article on AI Hallucination has adopted the definition in our paper! 📖 AI Hallucination Wiki: ) 📝 Our paper: 👍🏼 Shoutout to my collaborators! #GenAI

3

11

Aman Chadha

@i_amanchadha

8 months

🧠 A Detailed Overview of Vision-Language Models (VLMs) • ✅An overview of VLMs and ~40 popular VLM models for Generation and Understanding such as GPT-4V, LLaVA, Frozen, Flamingo, PaLM-E, MiniGPT, etc. #LLMs #ArtificialIntelligence #MachineLearning #AI

0

2

10

Aman Chadha

@i_amanchadha

9 months

📝 All you need to know about LLM Prompt Engineering • ⚡️ This primer presents 30+ flavors of Prompt Engineering along with their use-cases. #ArtificialIntelligence #MachineLearning #AI #ChatGPT

1

11

Aman Chadha

@i_amanchadha

6 months

📺The best #AI courses from Stanford, CMU, MIT, etc. (+ YouTube playlists) 🔗 🔹Stanford: CS221, CS229, CS230, CS231n, CS224n, CS224w, CS234, CS330, CS25 🔹CMU: CS 11-711, CS 11-747, CS 11-737, CS 11-777, CS 11-785 🔹MIT: 6.S191, 6.S094, 6.S192 🔹UCL: M050

1

7

13

Aman Chadha

@i_amanchadha

2 years

📝 Top papers in Computer Vision, NLP, Speech, Multimodal AI, and Core ML 👉🏼 I’ve put together a summary of key papers in Computer Vision, NLP, and Speech and segregated them into 1️⃣ need-to-know and 2️⃣ good-to-know. What’s…

2

9

Aman Chadha

@i_amanchadha

5 months

🤖 Policy Optimization Primer (𝐑𝐋𝐇𝐅, 𝐑𝐋𝐀𝐈𝐅, 𝐃𝐏𝐎, 𝐊𝐓𝐎, 𝐆𝐏𝐎, 𝐂𝐏𝐎, 𝐈𝐏𝐎, 𝐈𝐂𝐃𝐏𝐎, 𝐑𝐏𝐎, 𝐔𝐑𝐈𝐀𝐋, 𝐎𝐑𝐏𝐎, 𝐬𝐃𝐏𝐎, 𝐑𝐒-𝐃𝐏𝐎, 𝐌𝐃𝐏𝐎, 𝐒𝐢𝐦𝐏𝐎, 𝐅𝐚𝐜𝐭-𝐑𝐋𝐇𝐅, 𝐃𝐢𝐟𝐟𝐮𝐬𝐢𝐨𝐧-𝐃𝐏𝐎) 🔗 - Reinforcement Learning

1

4

10

Aman Chadha

@i_amanchadha

9 months

📣 I'm #hiring for full-time Generative AI roles in Amazon AWS. We're seeking exceptional talent interested in developing the next generation of AI-powered solutions. If this is you, get in touch! #ArtificialInteligence #MachineLearning #AI #Jobs

0

2

9

Aman Chadha

@i_amanchadha

10 months

🎖️Recommender Systems' Architectures: ➡️ This primer explores the most popular neural architectures used in recommender systems. 🔷 Wide and Deep, DeepFM, NCF, DCN, AutoInt, DCN V2, DHEN, GDCN #RecSys #ArtificialIntelligence #MachineLearning #AI

0

3

9

Aman Chadha

@i_amanchadha

6 months

📝 Recommender Systems in Production: Case Studies 🔹Synopses of the inner-workings of some of the most popular recommendation system platforms: - @tiktok_us 's Monolith Recommender System by Zhuoran Liu et al. - @netflix 's Recommender System by Carlos

0

1

9

Aman Chadha

@i_amanchadha

9 months

📝Announcing our @eaclmeeting 2024 papers 🔹On the Relationship between Sentence Analogy Identification and Sentence Structure Encoding in LLMs ➡️ 🔹Generative Data Augmentation using LLMs Improves Distributional Robustness in QA ➡️

0

9

Aman Chadha

@i_amanchadha

9 months

📺 The best AI courses from Stanford, CMU, MIT, DeepMind, UCL, NYU, UCF, UWaterloo, etc. (with YouTube playlists) |

0

3

8

Aman Chadha

@i_amanchadha

9 months

📈 Graph Neural Networks (GNNs) Primers: Overview, Applications in RecSys and NLP, and Case Studies ➡️ Overview of GNNs in RecSys and NLP: ➡️ GNN-based RecSys with Case-Studies from Snap, Meta, Google, Uber, Pinterest: #AI

0

9

Aman Chadha

@i_amanchadha

10 months

🧠 A Detailed Overview of Vision-Language Models (VLMs) | ✅ An overview of VLMs and ~40 popular VLM models for Generation and Understanding such as GPT-4V, LLaVA, Frozen, Flamingo, PaLM-E, MiniGPT, etc. #LLMs #ArtificialIntelligence #MachineLearning #AI

0

1

9

Aman Chadha

@i_amanchadha

8 months

🧠All you need to know about LLMs • ➡️Embeddings ➡️LLM Training Steps ➡️Context Length Scaling ➡️Traditional v/s Vector DBs ➡️Knowledge-Augmenting LLMs (RAG, Prompting, Finetuning) ➡️Popular LLMs (+Indic/Code LLMs) ➡️Leaderboards ➡️Frameworks #AI #LLMs

0

1

8

Aman Chadha

@i_amanchadha

10 months

📝 Announcing our new paper surveying 32 methods for mitigating hallucinations in LLMs 🔷 A Comprehensive Survey of Hallucination Mitigation Techniques in Large Language Models ➡️ #LLMs #ArtificialIntelligence #MachineLearning #AI

0

2

8

Aman Chadha

@i_amanchadha

6 months

📖All you need to know about LLMs: ➡️Embeddings ➡️LLM Training Steps ➡️Context Length Scaling ➡️Vector DBs ➡️Knowledge Augmenting LLMs (RAG, Prompting, Finetuning) ➡️50+ Popular LLMs (+Code / Indic / Medical LLMs) ➡️Leaderboards ➡️Frameworks #GenAI #LLMs

1

0

8

Aman Chadha

@i_amanchadha

4 months

📝 Announcing our paper that investigates biases in LLMs used for hate speech detection (focusing on gender, race, religion, and disability biases) and proposes mitigation strategies ➡️ 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐨𝐟 𝐀𝐧𝐧𝐨𝐭𝐚𝐭𝐨𝐫 𝐁𝐢𝐚𝐬: We demonstrate the presence of gender,

2

0

8

Aman Chadha

@i_amanchadha

10 months

🧠 All you need to know about Vision-Language Models (VLMs) | ➡️ A primer that offers an overview of VLMs and ~25 popular VLM architectures such as GPT-4V, LLaVA, Frozen, Flamingo, PaLM-E, MiniGPT, etc. #ArtificialIntelligence #MachineLearning #AI

0

1

8

Aman Chadha

@i_amanchadha

7 months

📝 Announcing our new paper with @UTAustin that explores socioeconomic biases in LLMs, revealing their lack of empathy towards the socioeconomically underprivileged ➡️Paper: ➡️Code: #GenAI #ArtificialIntelligence #AI #LLM #LLMs

0

1

8

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper that unveils the lack of cultural awareness in Vision-Language Models (VLMs)! ➡️ 𝐂𝐮𝐥𝐭𝐮𝐫𝐚𝐥 𝐀𝐰𝐚𝐫𝐞𝐧𝐞𝐬𝐬 𝐒𝐜𝐨𝐫𝐞 (𝐂𝐀𝐒): A novel metric proposed to measure the inclusion of culturally relevant information in image captions generated by

0

1

8

Aman Chadha

@i_amanchadha

3 months

📝 Announcing our ACL 2024 papers being presented next week 🔹 "𝐌𝐞𝐦𝐞𝐆𝐮𝐚𝐫𝐝: 𝐀𝐧 𝐋𝐋𝐌 𝐚𝐧𝐝 𝐕𝐋𝐌-𝐛𝐚𝐬𝐞𝐝 𝐅𝐫𝐚𝐦𝐞𝐰𝐨𝐫𝐤 𝐟𝐨𝐫 𝐀𝐝𝐯𝐚𝐧𝐜𝐢𝐧𝐠 𝐂𝐨𝐧𝐭𝐞𝐧𝐭 𝐌𝐨𝐝𝐞𝐫𝐚𝐭𝐢𝐨𝐧 𝐯𝐢𝐚 𝐌𝐞𝐦𝐞 𝐈𝐧𝐭𝐞𝐫𝐯𝐞𝐧𝐭𝐢𝐨𝐧" 🔹 Proposes MemeGuard, a novel

1

0

7

Aman Chadha

@i_amanchadha

6 months

🤖 Popular #LLMs (65+ Models) • ➡️ Foundation LLMs ➡️ Medical LLMs ➡️ Indic LLMs ➡️ Code LLMs #GenAI

1

3

7

Aman Chadha

@i_amanchadha

2 years

📚 Notes and Videos on Multimodal Machine Learning - The world surrounding us involves multiple modalities – we see objects, hear sounds, feel texture, smell odors, and so on. - In order for AI to make progress in understanding the world around us, it…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #deeplearning…

📚 Notes and Videos on Multimodal Machine Learning - The world surrounding us involves multiple modalities – we see objects, hear sounds, feel texture, smell…

www.linkedin.com

0

5

Aman Chadha

@i_amanchadha

10 months

🔹All you need to know about Large Language Models (LLMs): ➡️ Embeddings ➡️ LLM Training Steps ➡️ Context Length Scaling ➡️ Traditional v/s Vector DBs ➡️ Knowledge-Augmenting LLMs (RAG, Prompting, Finetuning) ➡️ Popular LLMs ➡️ Leaderboards ➡️ Frameworks

0

7

Aman Chadha

@i_amanchadha

10 months

📝 Announcing our @ECIR2024 paper with IIT Patna: 🔹MedSumm: A Multimodal Approach to Summarizing Code-Mixed Hindi-English Clinical Queries ➡️ ➡️ We introduce the novel task of Multimodal Medical Codemixed Question Summarization (+dataset,models) @IITPAT

1

0

7

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper that offers a comprehensive overview (surveying 100+ papers) of the Indic AI landscape ➡️ 𝐓𝐚𝐱𝐨𝐧𝐨𝐦𝐲 𝐚𝐧𝐝 𝐎𝐯𝐞𝐫𝐯𝐢𝐞𝐰: We survey the current state-of-the-art in Indic AI by covering LLMs, Corpora, Benchmarks and Evaluation, Techniques, Tools,

1

2

7

Aman Chadha

@i_amanchadha

6 months

🤖 Popular Multimodal LLMs (50+ Models) • ➡️ Text-Image VLMs for Generation ➡️ Text-Image VLMs for Understanding ➡️ Text-Video VLMs for Generation ➡️ Text-Video VLMs for Understanding ➡️ Any-to-Any VLMs #GenAI

0

3

7

Aman Chadha

@i_amanchadha

6 months

📝New paper w/ @USC : Comparative Study of Parameter Efficient Fine-Tuning (PEFT) techniques ✅We analyze the following PEFT techniques across commonsense reasoning, video-text understanding & medical imaging - Adapters - Prefix Tuning - Soft Prompt Tuning - BitFit - LoRA - LoReFT

1

7

Aman Chadha

@i_amanchadha

6 months

📈LLM/VLM Leaderboards • 🔹Open #LLM Leaderboard 🔹Chatbot Arena Leaderboard 🔹MTEB Leaderboard 🔹Open Medical-LLM Leaderboard 🔹Big Code Models Leaderboard 🔹Open #VLM Leaderboard 🔹Hallucination Leaderboard 🔹LLM-Perf Leaderboard 🔹Artificial Analysis

0

3

7

Aman Chadha

@i_amanchadha

6 months

📝 All you need to know about LLM Prompt Engineering • ⚡️ Here's my primer covering 30+ flavors of Prompt Engineering along with their use-cases. #ArtificialIntelligence #MachineLearning #AI #ChatGPT

0

2

7

Aman Chadha

@i_amanchadha

4 months

📝 A Curated Set of NLP Primers: Attention, Autoregressive vs. Autoencoder Models, Transformers, BERT 🔹 Attention: - Overview (The Attention Mechanism, The Bottleneck Problem, The Context Vector Bottleneck, How Attention Solves the Bottleneck Problem) -

0

2

8

Aman Chadha

@i_amanchadha

4 months

🔤 The A-to-Z of Word Embeddings • • Word embeddings are a crucial aspect of computational linguistics in NLP, providing a means for computers to interpret and process human language via nuanced language representation. • Grounded in distributional

0

1

6

Aman Chadha

@i_amanchadha

2 years

I’m #hiring for both full-time and intern roles for Amazon Alexa AI. We're looking for exceptional AI scientists skilled at deep learning, who are eager to work on cutting-edge AI, and build the next generation of Amazon #alexa . If this is you, get in tou…

Aman Chadha on LinkedIn: #hiring #alexa #amazon #jobs #artificialintelligence #machinelearning #ai…...

I’m #hiring for both full-time and intern roles for Amazon Alexa AI. We're looking for exceptional AI scientists skilled at deep learning, who are eager to… | 238 comments on LinkedIn

www.linkedin.com

1

6

Aman Chadha

@i_amanchadha

10 months

📚 Curated #NLP Primers 🔹Tokenizer: 🔹Attention: 🔹Autoregressive vs. Autoencoders: 🔹Transformers: 🔹BERT: 🔹Token Sampling Methods:

0

5

Aman Chadha

@i_amanchadha

6 months

📖 A-to-Z of Parameter Efficient Fine-Tuning (PEFT) • 🔹Soft Prompt Tuning 🔹Prefix Tuning 🔹Hard Prompt Tuning 🔹Adapters 🔹LoRA 🔹QLoRA 🔹QA-LoRA 🔹ReLoRA 🔹DoRA 🔹LoftQ 🔹LongLoRA 🔹MultiLoRA 🔹LQ-LoRA 🔹LoRA-FA 🔹Tied-LoRA 🔹GLoRA #GenAI #LLMs

0

7

Aman Chadha

@i_amanchadha

2 years

📝 All you need to know about Transformers, GPT, and BERT 🔹 Transformers: - Mathematical background (Vectors, Matrix Multiplication, Dot Product, Masking, Sampling) - Attention (Additive/Multiplicative/Dot Product Attention, Self…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #nlp | 57 comments

📝 All you need to know about Transformers, GPT, and BERT 🔹 Transformers: http://transformer.aman.ai - Mathematical background (Vectors, Matrix… | 57 comments on LinkedIn

www.linkedin.com

0

2

6

Aman Chadha

@i_amanchadha

2 years

I’m #hiring for both full-time and intern roles for Amazon Alexa AI. We're looking for exceptional AI scientists skilled at deep learning, folks who are eager to work on cutting-edge AI and build the next generation of Amazon #alexa . If this is you, get i…

Aman Chadha on LinkedIn: #hiring #alexa #amazon #jobs #artificialintelligence #machinelearning #ai…...

I’m #hiring for both full-time and intern roles for Amazon Alexa AI. We're looking for exceptional AI scientists skilled at deep learning, folks who are eager… | 723 comments on LinkedIn

www.linkedin.com

2

1

5

Aman Chadha

@i_amanchadha

8 months

📝 Announcing our new paper that reviews the impact of LLMs in Recommender Systems 🔹 "Exploring the Impact of Large Language Models on Recommender Systems: An Extensive Review" 🔹 In collaboration with @SantaClaraUniv and @CarnegieMellon ➡️ #AI #LLM

0

6

Aman Chadha

@i_amanchadha

9 months

📝 Announcing our new paper proposing a new debiasing technique for LLMs 🔹 "From Prejudice to Parity: A New Approach to Debiasing Large Language Model Word Embeddings" 🔹 In collaboration with @IITGuwahati , @UTAustin , and @UofIllinois 🔹 PDF: #AI #LLM

0

6

Aman Chadha

@i_amanchadha

7 months

📚All the Math fundamentals for #AI (+Backprop primers) ➡️Math primer: - Linear Algebra - Differential Calculus - Probability Theory & Distributions ➡️Backprop primer: - Chain Rule - Derivatives of Standard Layers/Loss Functions

0

2

6

Aman Chadha

@i_amanchadha

11 months

📝 We presented 5 papers @ #EMNLP2023 last week: 🔹LLM Hallucination: Definitions, Quantification, & Remediations | 🔹AI-Generated Text Detection: AI Detectability Index | @emnlpmeeting #ArtificialIntelligence #MachineLearning #AI

1

0

6

Aman Chadha

@i_amanchadha

6 months

📝 Top Papers in Computer Vision, NLP, Speech, Multimodal AI, Core ML, RecSys, & Graph ML 🔗 👉🏼 I’ve put together a summary of key papers in #AI and segregated them into (i) need-to-know and (ii) good-to-know. 🔹 Vision - Image Classification (CNN

1

0

6

Aman Chadha

@i_amanchadha

1 year

Our work at ACL 2023! Great to see our work piqued @chrmanning 's interest :) FACTIFY-5WQA: 5W Aspect-based Fact Verification through Question Answering Paper: #NLP #NLProc #ACL #ACL2023NLP

FACTIFY-5WQA: 5W Aspect-based Fact Verification through Question Answering

Automatic fact verification has received significant attention recently. Contemporary automatic fact-checking systems focus on estimating truthfulness using numerical scores which are not...

arxiv.org

Anku Rani

@anku__rani

1 year

Presented factify 5WQA, a small step towards building automatic and explainable fact verification system at ACL. Thanks @chrmanning for stopping by. It was great interacting with you. cc- @amit_p P.S. - I am open for PhD positions for fall’24. DM to chat. #NLProc #ACL2023

0

2

32

0

2

6

Aman Chadha

@i_amanchadha

2 years

📝 Recommender Systems in Production 🔹 Synopses and case studies of the inner-workings of some of the most popular recommendation system platforms: - TikTok's Monolith Recommender System by Zhuoran Liu et al. - Netflix's Recomme…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #nlp | 13 comments

📝 Recommender Systems in Production: Case Studies 🔹 Synopses of the inner-workings of some of the most popular recommendation system platforms:… | 13 comments on LinkedIn

www.linkedin.com

0

1

6

Aman Chadha

@i_amanchadha

8 months

📝Announcing our new paper with @NUSingapore and @Tsinghua_Uni that explores how inducing personalities in LLMs affects their Theory-of-Mind reasoning 🔹"PHAnToM: Personality Has An Effect on Theory-of-Mind Reasoning in Large Language Models" 🔗 #AI #LLMs

0

5

Aman Chadha

@i_amanchadha

8 months

📝 Announcing our new survey paper on "AI-generated Text Forensic Systems: Detection, Attribution, and Characterization" 🔹 In collaboration with @ASU 🔹 PDF: #ArtificialInteligence #AI #MachineLearning #LLMs

0

1

5

Aman Chadha

@i_amanchadha

7 months

📝Announcing our new paper with @UW on Claim Verification using LLMs + Knowledge Graphs (KGs) 🔹Framework that verifies claims against a trusted KG with explanations & attribution ✅Allows for pinpointing inaccuracies like #LLM hallucinations 🔗 #GenAI

0

6

Aman Chadha

@i_amanchadha

6 months

📚 Curated list of Books, Blogs, and YouTube Channels for AI and Data Science - I’ve curated a list of books, blogs and YouTube channels based on what has been meaningful for my career growth over the past several years. - These contain a broad range of domains in AI including

0

2

5

Aman Chadha

@i_amanchadha

10 months

(2/2)🔹 EMNLP 2023 Paper: "The Troubling Emergence of Hallucination in Large Language Models - An Extensive Definition, Quantification, and Prescriptive Remediations" ➡️ #GPT #LLMs #hallucinate #AI

The Troubling Emergence of Hallucination in Large Language Models - An Extensive Definition,...

Vipula Rawte, Swagata Chakraborty, Agnibh Pathak, Anubhav Sarkar, S.M Towhidul Islam Tonmoy, Aman Chadha, Amit Sheth, Amitava Das. Proceedings of the 2023 Conference on Empirical Methods in Natural...

aclanthology.org

0

2

Aman Chadha

@i_amanchadha

2 years

📚 Curated list of Books, Blogs, Course Notes, Newsletters, and YouTube Channels for AI and Data Science 👉🏼 I’ve curated a list of books, blogs and YouTube channels based on what has contributed to my career growth over the past few years. These contai…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #transformers… | 15...

📚 Curated list of Books, Blogs, Course Notes, Newsletters, and YouTube Channels for AI and Data Science 👉🏼 I’ve curated a list of books, blogs and YouTube… | 15 comments on LinkedIn

www.linkedin.com

0

3

Aman Chadha

@i_amanchadha

3 months

📖 Encoder vs. Decoder vs. Encoder-Decoder Models • - Among self-supervised representation learning objectives, encoder-based and decoder-based (i.e., autoregressive) language modeling have been the two most successful pretraining objectives. These two

0

1

5

Aman Chadha

@i_amanchadha

3 months

👷🏽 Machine Learning Infrastructure • - Infrastructure is the backbone of modern data-driven applications. - Here's my primer that covers the essential components that make up a successful ML project, from data ingestion to model deployment, and the major

0

1

5

Aman Chadha

@i_amanchadha

10 months

🔤 Primer on Word Embeddings | 🔹 Overview & Motivation 🔹 Conceptual Framework 🔹 Word Embedding Techniques (Count-based, Co-occurence-based, Contextualized) ➡️ TF-IDF ➡️ BM25 ➡️ Word2Vec ➡️ GloVe ➡️ fastText ➡️ BERT #AI #ML #NLP #LLMs

0

5

Aman Chadha

@i_amanchadha

5 months

📖 The A-to-Z of Parameter Efficient Fine-Tuning (PEFT) • - PEFT methods are crucial for foundation models such as LLMs because they allow for adapting these large models to specific tasks without needing to update all the parameters. This reduces

0

2

5

Aman Chadha

@i_amanchadha

4 months

📖 The A-to-Z of Distributed Training Parallelism • - Distributed training parallelism is crucial for efficiently training large-scale deep learning models that require extensive computational resources. This approach leverages multiple GPUs or machines

0

2

5

Aman Chadha

@i_amanchadha

6 months

🤖 LLM Alignment Primer • ✅ RLHF, RLAIF, DPO, KTO, GPO, CPO, IPO, ICDPO, ORPO, sDPO, RS-DPO, Diffusion-DPO - Reinforcement Learning with Human Feedback (RLHF), used to align LLM behavior with human preferences, has been pivotal for accurate and

0

7

Aman Chadha

@i_amanchadha

10 months

📚 All the Math fundamentals for AI 🔹 Math primer: - Linear Algebra & Differential Calculus - Probability Theory 🔹 Backprop primer: - Chain Rule - Derivatives of Layers and Loss Functions #ArtificialInteligence

0

4

Aman Chadha

@i_amanchadha

2 years

📝 Top papers in Computer Vision, NLP, Speech, and Core ML 👉🏼 I’ve put together a summary of key papers in Computer Vision, NLP, and Speech and segregated them into 1️⃣ need-to-know, and 2️⃣ good-to-know. What’s included (*not…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #researchscientist… | 42...

📝 Top papers in Computer Vision, NLP, Speech, and Core ML 👉🏼 http://papers.aman.ai I’ve put together a summary of key papers in Computer Vision, NLP, and… | 42 comments on LinkedIn

www.linkedin.com

0

3

4

Aman Chadha

@i_amanchadha

2 years

📚 All the Math fundamentals for AI 👉🏼 Don’t let the math behind AI concepts hinder you from understanding what goes on under-the-hood as you train your neural network. Think in terms of first principles to develop an intuition of what’s going on behin…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #deeplearning #research…...

📚 All the Math fundamentals for AI 👉🏼 Don’t let the math behind AI concepts hinder you from understanding what goes on under-the-hood as you train your… | 34 comments on LinkedIn

www.linkedin.com

0

4

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our ACL 2024 ( @aclmeeting ) paper on MemeGuard, a novel framework that leverages LLMs and VLMs for generating interventions to counteract the toxicity in cyberbullying memes ➡️ 𝐍𝐨𝐯𝐞𝐥 𝐓𝐚𝐬𝐤 𝐚𝐧𝐝 𝐃𝐚𝐭𝐚𝐬𝐞𝐭: We introduce the task of meme intervention and

0

3

Aman Chadha

@i_amanchadha

1 year

📝 Announcing our new paper at ICASSP 2023 next month! 🔹 Paper: “I See What You Hear: A Vision-inspired Method To Localize Words” with Mohammad Samragh Razlighi, Arnav Kundu, Hu Ting-Yao, Minsik Cho, Ashish Shrivastava, @OncelTuzel , and Devang Naik 🔹…

0

1

4

Aman Chadha

@i_amanchadha

10 months

📚 Curated list of Books, Blogs, & YouTube Channels for AI 🔹 Read List: - Stanford, MIT, & CMU course notes, blogs, books, etc. 🔹 Watch List: - Course videos & YouTube tutorials #ArtificialIntelligence #MachineLearning #LLMs

0

4

Aman Chadha

@i_amanchadha

4 years

Announcing my new #AI paper on Video Super Resolution that builds on Recurrent Back Projection Networks using GANs with a four-fold loss. We’re #1 on the Video Super Resolution leaderboard! 🙂 #DeepLearning #ArtificialIntelligence #NeuralNetworks

0

2

4

Aman Chadha

@i_amanchadha

5 months

📝 Tutorial: Hallucination in LLMs ✅ Our tutorial at @LrecColing 2024 last week offered an introduction to the issue of hallucination, presented a taxonomy of categories, and covered the top hallucination detection and mitigation methods. ✅ We also covered the next big

0

4

Aman Chadha

@i_amanchadha

9 months

📝Announcing our @ieeeICASSP 2024 paper: "Efficient Post-Model Training Speaker Embedding Space Alignment for Asymmetric Model Deployment" ✅A novel method for aligning speaker embedding spaces in asymmetric speaker identification systems. ➡️ #AI #speech

0

4

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper that (i) demonstrates that SLMs can effectively compete with – and sometimes outperform – frontier LLMs such as GPT-4 when appropriately selected and prompted, and (ii) proposes a framework for selecting the best model and prompt style based on the

0

1

4

Aman Chadha

@i_amanchadha

5 months

📝 Announcing our paper surveying Accelerated Generation Techniques in LLMs ➡️ We categorize 50+ accelerated generation techniques in LLMs into: - Speculative Decoding - Early Exiting - Non-Autoregressive Methods ➡️ Speculative decoding explores multiple candidate outputs

1

0

3

Aman Chadha

@i_amanchadha

2 years

🧠 Curated set of NLP Primers: A one-stop shop! 👉🏼 Primers for Attention, Autoregressive vs. Autoencoder Models, Transformers, BERT, BigBird 🔹 Attention: - Classic Sequence-to-Sequence Model - Sequence-to-Sequence Model with A…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #deeplearning #research…...

🧠 A curated set of NLP Primers is all you need! 👉🏼 Primers for the common NLP topics: Attention, Autoregressive vs. Autoencoder Models, Transformers, BERT… | 21 comments on LinkedIn

www.linkedin.com

0

4

Aman Chadha

@i_amanchadha

1 year

🤓 Primers for AI and Data Science: Python, PyTorch, TensorFlow, NumPy, Pandas, Matplotlib 👉🏼 Looking to kickstart a career in AI? Here are 6 primers covering some of the pillars of AI and Data Science: Python, PyTorch, TensorFlow, NumPy, Pandas, and M…

0

2

4

Aman Chadha

@i_amanchadha

2 years

🤖 Recommender Systems (RecSys) Primer: The why (motivation), what (concepts under-the-hood), and how (code deep-dive) 👉🏼 TL;DR: - Recommender models are an integral part of product offerings from every major company, with ofte…

Aman Chadha on LinkedIn: #alexa #recsys #artificialintelligence #machinelearning #ai #ml…

🤖 Recommender Systems (RecSys) Primer: The why (motivation), what (concepts under-the-hood), and how (code deep-dive) 👉🏼 TL;DR: http://recsys.aman.ai -…

www.linkedin.com

0

1

4

Aman Chadha

@i_amanchadha

1 year

🤖 Large Language Model (LLM) Primers | ChatGPT, Prompt Engineering, RLHF With the advent of ChatGPT, LLMs have been the talk of the town! We’ve recently seen a bunch of extraordinary advancements in the NLP space including GPT-4, LLaMA, Toolformer, RLHF…

Aman Chadha on LinkedIn: #artificialintelligence #machinelearning #ai #ml #nlp | 23 comments

🤖 Large Language Model (LLM) Primers | ChatGPT, Prompt Engineering, RLHF With the advent of ChatGPT, LLMs have been the talk of the town! We’ve recently seen… | 23 comments on LinkedIn

www.linkedin.com

0

4