AI Named This Show is a weekly AI-focused tech show. Join longtime friends and tech media veterans Tasia Custode and Tristan Jutras as they dive into the AI abyss, unraveling the complexities of artificial intelligence. They cover everything from groundbreaking AI news to the practical applications — and societal implications — of large language models, machine learning, deep learning, generative AI and more. Hosted on Acast. See acast.com/privacy for more information.
…
continue reading
Are you a critical thinker eager to dive into AI? Welcome to The Generative AI Podcast: Super Prompt. We leverage the latest industry developments to build foundational knowledge in AI. Join me, Tony Wan, an ex Silicon Valley executive, as we 'unhype the hype' of AI via illuminating conversations with top engineers and entrepreneurs, complemented by in-depth solo episodes. Our goal? To make it almost unnecessary to send a cybernetic organism back in time to fix things. Tailored for the techn ...
…
continue reading
A daily update on the latest AI Research Papers. We provide a high level overview of a handful of papers each day and will link all papers in the description for further reading. This podcast is created entirely with AI by PocketPod. Head over to https://pocketpod.app to learn more.
…
continue reading
Welcome to an exciting new season of the podcast Your Career: Choice or Chance? - as we dive into the ever-evolving world of GenAI in the Workplace and explore the latest trends, experiences, and career journeys shaping the future of work as AI is increasingly ingrained in it. Each episode provides fresh insights, addressing the transformative influence of GenAI in shaping the workforce of tomorrow, making it a must-listen for anyone interested in staying ahead in the ever-evolving world of ...
…
continue reading
Tune in as we dissect recent AI news, explore cutting-edge innovations, and sit down with influential voices shaping the future of AI. Whether you're a seasoned expert or just dipping your toes into the AI waters, our podcast is your go-to resource for staying informed and inspired. #IntelAI @IntelAI
…
continue reading
Your AI Roadmap the podcast is on a mission to decrease fluffy HYPE and talk to the people actually building AI. Anyone can build in AI. Including you. Whether you’re terrified or excited, there’s been no better time than today to dive in! Now is the time to be curious and future-proof your career and ... ultimately your income. This podcast isn't about white dudes patting themselves on the back, this is about you and me and ALL the paths into cool projects around the world! What's next on y ...
…
continue reading
"On AI" is an innovative podcast uniquely tailored for creators diving into the fascinating world of generative AI. Generative AI is accelerating artistic expressions resulting in the development of completely new genres, transforming art, design, film, music, storytelling and immersive multimodal experiences. It's also transforming the way we experience the world. From fashion, gaming, robotics and all the pop culture in between. In each episode, we examine the transformative power of artif ...
…
continue reading
Weekly talks and fireside chats about everything that has to do with the new space emerging around DevOps for Machine Learning aka MLOps aka Machine Learning Operations.
…
continue reading
Looking to explore the intersection of AI and journalism? Influential thought leaders in the industry join data scientist and media entrepreneur, Nikita Roy, each week to explore what's next with AI and its implications for the media landscape. In each episode, industry experts discuss how automated newsrooms have the potential to change journalism and uncover opportunities to optimize workflows and increase efficiency without compromising journalistic integrity. Hosted on Acast. See acast.c ...
…
continue reading
EdgeCortix subject matter experts discuss edge AI processors, AI software frameworks, and AI industry trends.
…
continue reading
The leading podcast on how to build a successful open source company. Learn from the founders of HashiCorp, Chronosphere, Vercel, MongoDB, DBT, mobile.dev and more!
…
continue reading
A weekly show discussing the latest developments in the European technology ecosystem and featuring interviews with some of the most interesting people in the industry.
…
continue reading
Running out of time to catch up with new arXiv papers? We take the most impactful papers and present them as convenient podcasts. If you're a visual learner, we offer these papers in an engaging video format. Our service fills the gap between overly brief paper summaries and time-consuming full paper reads. You gain academic insights in a time-efficient, digestible format. Code behind this work: https://github.com/imelnyk/ArxivPapers Support this podcast: https://podcasters.spotify.com/pod/s ...
…
continue reading
I make videos about machine learning research papers, programming, and issues of the AI community, and the broader impact of AI in society. Twitter: https://twitter.com/ykilcher Discord: https://discord.gg/4H8xxDF If you want to support me, the best thing to do is to share out the content :) If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this): SubscribeStar (preferred to Patreon): https://www.subscribestar.com/yannickilcher Patre ...
…
continue reading
eDiscovery Data Points are selected articles published on the ComplexDiscovery blog and shared to update legal, information technology, and business professionals on the art and science of data discovery and legal discovery.
…
continue reading
Seed to Harvest, hosted by Paige Finn Doherty, highlights stories, frameworks & tactics from a diverse array of investors, founders, and creators. If you're interested in investing or building a business, this show is for you!
…
continue reading
The AR Show dives deep into the emerging world of Augmented Reality with a focus on the underlying technologies and uses of Smartglasses, and the people behind them. I talk with entrepreneurs, executives, investors and early adopters to extract insights that will both inform and inspire you. In each episode, I explore the approaches, challenges, and progress behind the products and companies. I also extract the lessons learned and insightful advice from each guest. Equal parts technology, pr ...
…
continue reading
1
From satellite tracking and space regulation to multimodal AI, and more
38:54
38:54
Play later
Play later
Lists
Like
Liked
38:54
Welcome to the new episode of the TNW Podcast — the show where we discuss the latest developments in the European technology ecosystem and feature interviews with some of the most interesting people in the industry. In today’s episode, we're featuring two interviews recorded at TNW Conference 2024 ago with two amazing women working in very differen…
…
continue reading
1
OMG-LLaVA: Unifying Vision and Language Understanding, Step-DPO for LLMs Mathematical Reasoning, MUMU's Multimodal Image Generation
12:15
12:15
Play later
Play later
Lists
Like
Liked
12:15
OMG-LLaVA: Bridging Image-level, Object-level, Pixel-level Reasoning and UnderstandingStep-DPO: Step-wise Preference Optimization for Long-chain Reasoning of LLMsMUMU: Bootstrapping Multimodal Image Generation from Text-to-Image DataSimulating Classroom Education with LLM-Empowered AgentsSeaKR: Self-aware Knowledge Retrieval for Adaptive Retrieval …
…
continue reading
1
Accelerating Multimodal AI // Ethan Rosenthal // #242
54:57
54:57
Play later
Play later
Lists
Like
Liked
54:57
Join us at our first in-person conference on June 25 all about AI Quality: https://www.aiqualityconference.com/Accelerating Multimodal AI // MLOps podcast #241 with Ethan Rosenthal, Member of Technical Staff of Runway.Huge thank you to AWS for sponsoring this episode. AWS - https://aws.amazon.com/// AbstractWe’re still trying to figure out systems …
…
continue reading
1
Multimodal AI, Self-Supervised Learning, Counterfactual Reasoning, and AI Agents with Vasudev Lal
37:28
37:28
Play later
Play later
Lists
Like
Liked
37:28
Discover the cutting-edge advancements in artificial intelligence with Vasudev Lal, Principal AI Research Scientist at Intel. This episode delves into the benefits of multimodal AI and the enhanced validity achieved through self-supervised learning. Vasudev also explores the applications of counterfactual reasoning in AI and the efficiency gains fr…
…
continue reading
1
Meta GenAI Infra Blog Review // Special MLOps Podcast
38:53
38:53
Play later
Play later
Lists
Like
Liked
38:53
Meta GenAI Infra Blog Review // Special MLOps Podcast episode by Demetrios.// AbstractDemetrios explores Meta's innovative infrastructure for large-scale AI operations, highlighting three blog posts on training large language models, maintaining AI capacity, and building Meta's GenAI infrastructure. The discussion reveals Meta's handling of hundred…
…
continue reading
1
Persona-Driven Data Synthesis, Enhancing Medical MLLMs, Robot Learning, Knowledge Distillation in LLMs, Text to 3D Gaussian Revolution
11:24
11:24
Play later
Play later
Lists
Like
Liked
11:24
Scaling Synthetic Data Creation with 1,000,000,000 PersonasHuatuoGPT-Vision, Towards Injecting Medical Visual Knowledge into Multimodal LLMs at ScaleLLaRA: Supercharging Robot Learning Data for Vision-Language PolicyDirect Preference Knowledge Distillation for Large Language ModelsGaussianDreamerPro: Text to Manipulable 3D Gaussians with Highly Enh…
…
continue reading
1
[QA] Searching for Best Practices in Retrieval-Augmented Generation
10:22
10:22
Play later
Play later
Lists
Like
Liked
10:22
The paper explores optimal practices for retrieval-augmented generation (RAG) to improve response quality and efficiency, suggesting strategies and demonstrating benefits of multimodal retrieval techniques. https://arxiv.org/abs//2407.01219 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
1
Searching for Best Practices in Retrieval-Augmented Generation
22:28
22:28
Play later
Play later
Lists
Like
Liked
22:28
The paper explores optimal practices for retrieval-augmented generation (RAG) to improve response quality and efficiency, suggesting strategies and demonstrating benefits of multimodal retrieval techniques. https://arxiv.org/abs//2407.01219 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
1
Generative AI for Enterprise with Noelle Russell of AI Leadership Institute
45:42
45:42
Play later
Play later
Lists
Like
Liked
45:42
Noelle Russell, founder of the AI Leadership Institute, shares her experience in AI, highlighting responsible AI, challenges in implementation, and the potential of generative AI in customer service. She emphasizes collaboration, flexibility, and aligning values with partners. Russell discusses harnessing team knowledge with LLMs, the role of certi…
…
continue reading
Analysis of current AI agent benchmarks reveals shortcomings in evaluation practices, focusing on accuracy over cost, leading to complex agents. Proposed solutions aim to optimize cost and accuracy jointly. https://arxiv.org/abs//2407.01502 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
Analysis of current AI agent benchmarks reveals shortcomings in evaluation practices, focusing on accuracy over cost, leading to complex agents. Proposed solutions aim to optimize cost and accuracy jointly. https://arxiv.org/abs//2407.01502 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
1
E140: Accelerating Enterprise AI Adoption with Better Agentic Workflows
35:13
35:13
Play later
Play later
Lists
Like
Liked
35:13
Mark Huang is Co-Founder of Gradient, the platform for enterprise agentic automation. Gradient recently open sourced their 4M context window finetune of Llama-3, which is the longest context window available today. Gradient has raised $10M from investors including Wing VC, Mango Capital, and Tokyo Black. In this episode, we dig into enterprise read…
…
continue reading
1
[QA] Token Erasure as a Footprint of Implicit Vocabulary Items in LLMs
10:36
10:36
Play later
Play later
Lists
Like
Liked
10:36
LLMs process text with tokens, but individual tokens may not relate to word meanings. This study explores how LLMs convert tokens into higher-level representations, revealing an "erasure" effect. https://arxiv.org/abs//2406.20086 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podca…
…
continue reading
1
Token Erasure as a Footprint of Implicit Vocabulary Items in LLMs
8:50
8:50
Play later
Play later
Lists
Like
Liked
8:50
LLMs process text with tokens, but individual tokens may not relate to word meanings. This study explores how LLMs convert tokens into higher-level representations, revealing an "erasure" effect. https://arxiv.org/abs//2406.20086 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podca…
…
continue reading
1
[QA] Scaling Synthetic Data Creation with 1,000,000,000 Personas
11:56
11:56
Play later
Play later
Lists
Like
Liked
11:56
The paper introduces Persona Hub, a collection of 1 billion diverse personas, to create diverse synthetic data at scale for various applications, showcasing its versatility and potential impact. https://arxiv.org/abs//2406.20094 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcas…
…
continue reading
1
Scaling Synthetic Data Creation with 1,000,000,000 Personas
16:08
16:08
Play later
Play later
Lists
Like
Liked
16:08
The paper introduces Persona Hub, a collection of 1 billion diverse personas, to create diverse synthetic data at scale for various applications, showcasing its versatility and potential impact. https://arxiv.org/abs//2406.20094 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcas…
…
continue reading
1
[QA] Is Programming by Example solved by LLMs?
9:45
9:45
Play later
Play later
Lists
Like
Liked
9:45
Large Language Models (LLMs) show promise in solving Programming-by-Examples (PBE) tasks but require fine-tuning for better performance, especially for out-of-distribution problems. https://arxiv.org/abs//2406.08316 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/…
…
continue reading
1
Is Programming by Example solved by LLMs?
16:57
16:57
Play later
Play later
Lists
Like
Liked
16:57
Large Language Models (LLMs) show promise in solving Programming-by-Examples (PBE) tasks but require fine-tuning for better performance, especially for out-of-distribution problems. https://arxiv.org/abs//2406.08316 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/…
…
continue reading
1
[QA] Can LLMs Learn by Teaching? A Preliminary Study
9:27
9:27
Play later
Play later
Lists
Like
Liked
9:27
The paper explores whether Language Models can learn by teaching (LbT) like humans, showing promising results in improving models through teaching methods. https://arxiv.org/abs//2406.14629 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id…
…
continue reading
1
Can LLMs Learn by Teaching? A Preliminary Study
18:28
18:28
Play later
Play later
Lists
Like
Liked
18:28
The paper explores whether Language Models can learn by teaching (LbT) like humans, showing promising results in improving models through teaching methods. https://arxiv.org/abs//2406.14629 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id…
…
continue reading
1
[QA] Found in the Middle: Calibrating Positional Attention Bias Improves Long Context Utilization
11:31
11:31
Play later
Play later
Lists
Like
Liked
11:31
Large language models struggle with capturing relevant information in the middle of input due to intrinsic attention bias. Mitigating bias with found-in-the-middle mechanism improves performance and RAG outcomes significantly. https://arxiv.org/abs//2406.16008 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_paper…
…
continue reading
1
Found in the Middle: Calibrating Positional Attention Bias Improves Long Context Utilization
15:38
15:38
Play later
Play later
Lists
Like
Liked
15:38
Large language models struggle with capturing relevant information in the middle of input due to intrinsic attention bias. Mitigating bias with found-in-the-middle mechanism improves performance and RAG outcomes significantly. https://arxiv.org/abs//2406.16008 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_paper…
…
continue reading
1
[QA] Computational Life: How Well-formed, Self-replicating Programs Emerge from Simple Interaction
8:38
8:38
Play later
Play later
Lists
Like
Liked
8:38
This paper explores the emergence of self-replicators on computational substrates, showing how they arise through random interactions and self-modification, leading to complex dynamics. https://arxiv.org/abs//2406.19108 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.…
…
continue reading
1
Computational Life: How Well-formed, Self-replicating Programs Emerge from Simple Interaction
21:19
21:19
Play later
Play later
Lists
Like
Liked
21:19
This paper explores the emergence of self-replicators on computational substrates, showing how they arise through random interactions and self-modification, leading to complex dynamics. https://arxiv.org/abs//2406.19108 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.…
…
continue reading
This week, Tristan and Tasia relive a tech journalist's chaotic adventure with Meta's Ray-Ban AI glasses in Montreal and uncover the significant security flaw found in Rabbit’s R1 AI gadget. Then we dive into the heated legal battles over AI-generated music, with the RIAA suing Suno and Udio for copyright infringement and YouTube negotiating music …
…
continue reading
1
AI Agents for Consumers // Shaun Wei // #244
57:26
57:26
Play later
Play later
Lists
Like
Liked
57:26
Sean Wei, the CEO and co-founder of RealChar, shares his journey from working in the autonomous vehicle industry to creating an open-source voice assistant project called Realchar, which eventually evolved into Rivia, a voice AI assistant focused on managing personal phone calls.The Future of AI and Consumer Empowerment // MLOps podcast #244 with S…
…
continue reading
1
FineWeb Datasets, YouDream's 3D Animals, PDE-Solving Breakthrough, Noise-Conditioned Perception Alignment, Language Models' Continual Learning
11:02
11:02
Play later
Play later
Lists
Like
Liked
11:02
The FineWeb Datasets: Decanting the Web for the Finest Text Data at ScaleYouDream: Generating Anatomically Controllable Consistent Text-to-3D AnimalsDiffusionPDE: Generative PDE-Solving Under Partial ObservationAligning Diffusion Models with Noise-Conditioned PerceptionUnlocking Continual Learning Abilities in Language Models…
…
continue reading
1
[QA] REVISION MATTERS: Generative Design Guided by Revision Edits
9:09
9:09
Play later
Play later
Lists
Like
Liked
9:09
Investigating how human designer revisions benefit a multimodal generative model for layout design, showing expert edits lead to strong design outcomes, emphasizing human guidance for iterative improvement. https://arxiv.org/abs//2406.18559 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
1
REVISION MATTERS: Generative Design Guided by Revision Edits
13:36
13:36
Play later
Play later
Lists
Like
Liked
13:36
Investigating how human designer revisions benefit a multimodal generative model for layout design, showing expert edits lead to strong design outcomes, emphasizing human guidance for iterative improvement. https://arxiv.org/abs//2406.18559 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: ht…
…
continue reading
1
[QA] The Remarkable Robustness of LLMs: Stages of Inference?
9:20
9:20
Play later
Play later
Lists
Like
Liked
9:20
The study explores Large Language Models' robustness by deleting and swapping layers, finding interventions retain 72-95% accuracy without fine-tuning, with more layers showing increased robustness. https://arxiv.org/abs//2406.19384 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://po…
…
continue reading
1
The Remarkable Robustness of LLMs: Stages of Inference?
15:11
15:11
Play later
Play later
Lists
Like
Liked
15:11
The study explores Large Language Models' robustness by deleting and swapping layers, finding interventions retain 72-95% accuracy without fine-tuning, with more layers showing increased robustness. https://arxiv.org/abs//2406.19384 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://po…
…
continue reading
1
E139: Taking on AWS with an Open Source Alternative
38:05
38:05
Play later
Play later
Lists
Like
Liked
38:05
Umur Cubukcu is Co-Founder of Ubicloud, the open source and portable cloud that can reduce cloud spend by 3–10x. Their project, also called ubicloud, has over 3K stars and provides elastic compute, block storage, virtual networking, managed Postgres, and IAM services. Ubicloud has raised $16M from investors including 500 Global and YC. In this epis…
…
continue reading
1
BigCodeBench Challenges, Cambrian-1 Leap, D-MERIT's Evaluation, Long Context Breakthrough in Vision
11:06
11:06
Play later
Play later
Lists
Like
Liked
11:06
DreamBench++: A Human-Aligned Benchmark for Personalized Image GenerationBigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex InstructionsCambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMsEvaluating D-MERIT of Partial-annotation on Information RetrievalLong Context Transfer from Language to Vision…
…
continue reading
1
[QA] Do LLMs dream of elephants (when told not to)? Latent concept association and associative memory in transformers
9:37
9:37
Play later
Play later
Lists
Like
Liked
9:37
Large Language Models can easily manipulate fact retrieval by changing contexts, behaving like associative memory models. Transformers use self-attention and value matrix for memory tasks. https://arxiv.org/abs//2406.18400 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.app…
…
continue reading
1
Do LLMs dream of elephants (when told not to)? Latent concept association and associative memory in transformers
16:26
16:26
Play later
Play later
Lists
Like
Liked
16:26
Large Language Models can easily manipulate fact retrieval by changing contexts, behaving like associative memory models. Transformers use self-attention and value matrix for memory tasks. https://arxiv.org/abs//2406.18400 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.app…
…
continue reading
1
[QA] Data curation via joint example selection further accelerates multimodal learning
8:30
8:30
Play later
Play later
Lists
Like
Liked
8:30
Jointly selecting batches of data improves learning in large-scale pretraining. Multimodal contrastive objectives reveal data dependencies, leading to faster training with reduced computational overhead. https://arxiv.org/abs//2406.17711 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https…
…
continue reading
1
Data curation via joint example selection further accelerates multimodal learning
13:49
13:49
Play later
Play later
Lists
Like
Liked
13:49
Jointly selecting batches of data improves learning in large-scale pretraining. Multimodal contrastive objectives reveal data dependencies, leading to faster training with reduced computational overhead. https://arxiv.org/abs//2406.17711 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https…
…
continue reading
1
Live from TNW 2024! Maria Amelie on fact-checking with AI; post-quantum cryptography; competitive Excel
37:58
37:58
Play later
Play later
Lists
Like
Liked
37:58
In today’s episode — recorded with live audience at TNW Conference 2024 — Linnea and Andrii talk about post-quantum cryptography, competitive Excel, and a few more things in between. The guest of the show is Maria Amelie, CEO and founder of Factiverse. The company has just raised €1mn in funding to further build its platform that helps researchers,…
…
continue reading
1
[QA] Large Language Models are Interpretable Learners
9:29
9:29
Play later
Play later
Lists
Like
Liked
9:29
Combining Large Language Models with symbolic programs creates interpretable and accurate decision rules, bridging the gap between expressiveness and interpretability in predictive models. https://arxiv.org/abs//2406.17224 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.app…
…
continue reading
1
Large Language Models are Interpretable Learners
20:30
20:30
Play later
Play later
Lists
Like
Liked
20:30
Combining Large Language Models with symbolic programs creates interpretable and accurate decision rules, bridging the gap between expressiveness and interpretability in predictive models. https://arxiv.org/abs//2406.17224 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.app…
…
continue reading
1
LongRAG Breakthrough, LLMs as Judges, Transformer Memory Insights, Video Library AI, Democratizing Art Styles
10:14
10:14
Play later
Play later
Lists
Like
Liked
10:14
LongRAG: Enhancing Retrieval-Augmented Generation with Long-context LLMsJudging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-JudgesComplexity of Symbolic Representation in Working Memory of Transformer Correlates with the Complexity of a TaskTowards Retrieval Augmented Generation over Large Video LibrariesStylebreeder: Exploring …
…
continue reading
1
[QA] Recite, Reconstruct, Recollect: Memorization in LMs as a Multifaceted Phenomenon
7:42
7:42
Play later
Play later
Lists
Like
Liked
7:42
Memorization in language models is complex and influenced by various factors. A taxonomy approach helps understand and predict memorization patterns. https://arxiv.org/abs//2406.17746 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id169247…
…
continue reading
1
Recite, Reconstruct, Recollect: Memorization in LMs as a Multifaceted Phenomenon
12:44
12:44
Play later
Play later
Lists
Like
Liked
12:44
Memorization in language models is complex and influenced by various factors. A taxonomy approach helps understand and predict memorization patterns. https://arxiv.org/abs//2406.17746 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/podcast/arxiv-papers/id169247…
…
continue reading
1
ML and AI as Distinct Control Systems in Heavy Industrial Settings // Richard Howes // #243
56:30
56:30
Play later
Play later
Lists
Like
Liked
56:30
Join us at our first in-person conference today all about AI Quality: https://www.aiqualityconference.com/ML and AI as Distinct Control Systems in Heavy Industrial Settings // MLOps podcast #243 with Richard Howes, CTO of Metaformed. Richard Howes is a dedicated engineer who is passionate about control systems whether it be embedded systems, indust…
…
continue reading
1
[QA] Adam-mini: Use Fewer Learning Rates To Gain More
7:57
7:57
Play later
Play later
Lists
Like
Liked
7:57
Adam-mini optimizer reduces memory footprint by using average learning rates within parameter blocks, achieving performance comparable to AdamW with significantly less memory. https://arxiv.org/abs//2406.16793 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/pod…
…
continue reading
1
Adam-mini: Use Fewer Learning Rates To Gain More
13:47
13:47
Play later
Play later
Lists
Like
Liked
13:47
Adam-mini optimizer reduces memory footprint by using average learning rates within parameter blocks, achieving performance comparable to AdamW with significantly less memory. https://arxiv.org/abs//2406.16793 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.com/us/pod…
…
continue reading
1
[QA] Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs
11:02
11:02
Play later
Play later
Lists
Like
Liked
11:02
SEPs offer a cost-effective method for detecting hallucinations in Large Language Models by approximating semantic entropy from hidden states, improving efficiency and generalization. https://arxiv.org/abs//2406.15927 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.co…
…
continue reading
1
Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs
14:49
14:49
Play later
Play later
Lists
Like
Liked
14:49
SEPs offer a cost-effective method for detecting hallucinations in Large Language Models by approximating semantic entropy from hidden states, improving efficiency and generalization. https://arxiv.org/abs//2406.15927 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Podcasts: https://podcasts.apple.co…
…
continue reading
1
Crafting Copilot's AI with Stephanie Blucker of Microsoft
37:54
37:54
Play later
Play later
Lists
Like
Liked
37:54
Stephanie Blucker, a senior content design manager at Microsoft, shares insights into her team's work on AI integrations across products like Word, Excel, PowerPoint, Outlook, Loop, Planner, & the M365 app. She explains the evolution of content design from UX writing to a more integrated role, emphasizing content designers' importance in creating u…
…
continue reading
1
[QA] Evaluating Numerical Reasoning in Text-to-Image Models
11:41
11:41
Play later
Play later
Lists
Like
Liked
11:41
Text-to-image models struggle with numerical reasoning tasks, showing limitations in generating exact numbers, understanding quantifiers, zero, and advanced concepts. GECKONUM benchmark is introduced for evaluation. https://arxiv.org/abs//2406.14774 YouTube: https://www.youtube.com/@ArxivPapers TikTok: https://www.tiktok.com/@arxiv_papers Apple Pod…
…
continue reading