For this week's paper read, we actually dive into our own research. We wanted to create a replicable, evolving dataset that can keep pace with model training so that you always know you're testing with data your model has never seen before. We also saw the prohibitively high cost of running LLM evals at scale, and have used our data to fine-tune a …

Valerie Abend, Accenture’s financial services cybersecurity lead, explains what banks get wrong about fending off AI-based threats and what they should do instead.

This week we talk about modern AI benchmarks, taking a close look at Google's recent Gemini 2.5 release and its performance on key evaluations, notably Humanity's Last Exam (HLE). In the session we covered Gemini 2.5's architecture, its advancements in reasoning and multimodality, and its impressive context window. We also talked about how benchmar…

Generative AI will remove toil from our day to day jobs, argues Lukic, who is managing director and senior partner at Boston Consulting Group.

We cover Anthropic’s groundbreaking Model Context Protocol (MCP). Though it was released in November 2024, we've been seeing a lot of hype around it lately, and thought it was well worth digging into. Learn how this open standard is revolutionizing AI by enabling seamless integration between LLMs and external data sources, fundamentally transformin…

“We want to put banks in the risk management driver’s seat,” says Sima Gandhi, co-founder of the Council for Fintech Ecosystem Standards, which has worked with a group of fintechs to create risk and compliance standards banks can use to evaluate their fintech partners.

Software code generation and knowledge management are two of the places the bank has begun using generative AI to improve efficiency.

This week, we're mixing things up a little bit. Instead of diving deep into a single research paper, we cover the biggest AI developments from the past few weeks. We break down key announcements, including: DeepSeek’s Big Launch Week: A look at FlashMLA (DeepSeek’s new approach to efficient inference) and DeepEP (their enhanced pretraining method).…

This week, we dive into DeepSeek. SallyAnn DeLucia, Product Manager at Arize, and Nick Luzio, a Solutions Engineer, break down key insights on a model that have dominating headlines for its significant breakthrough in inference speed over other models. What’s next for AI (and open source)? From training strategies to real-world performance, here’s …

Thomas Rudzewick, CEO of Maspeth Savings Bank worries about the growing crisis of low savings among millennials and Gen Z. He believes banks like his can help reverse this trend with financial literacy and innovative savings tools.

We talk to Google DeepMind Senior Research Scientist (and incoming Assistant Professor at Harvard), Yilun Du, about his latest paper "Multiagent Finetuning: Self Improvement with Diverse Reasoning Chains." This paper introduces a multiagent finetuning framework that enhances the performance and diversity of language models by employing a society of…

The bank, which is headquartered in Pasadena, had to quickly switch to remote work for many employees and come up with relief programs for customers whose homes and businesses were destroyed by fire.

Arvind Purushotham, head of Citi Ventures, shares where he and his team see opportunities and how they vet tech startups.

LLMs have typically been restricted to reason in the "language space," where chain-of-thought (CoT) is used to solve complex reasoning problems. But a new paper argues that language space may not always be the best for reasoning. In this paper read, we cover an exciting new technique from a team at Meta called Chain of Continuous Thought—also known…

A lack of resources is one common cause of AML penalties, says consultant Aaron Ansari.

We discuss a major survey of work and research on LLM-as-Judge from the last few years. "LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods" systematically examines the LLMs-as-Judge framework across five dimensions: functionality, methodology, applications, meta-evaluation, and limitations. This survey gives us a birds eye view…

Some banks are “punching above their weight,” according to Dan Latimore, chief research officer at The Financial Revolutionist. Here’s how they do it.

LLMs have revolutionized natural language processing, showcasing remarkable versatility and capabilities. But individual LLMs often exhibit distinct strengths and weaknesses, influenced by differences in their training corpora. This diversity poses a challenge: how can we maximize the efficiency and utility of LLMs? A new paper, "Merge, Ensemble, a…

Ian Mitchell, founder of The Noble, an organization that works with law enforcement and with banks to fight human crime trafficking, explains some of his group’s recent work and why banks need someone dedicated to human crime.

This week, we break down the “Agent-as-a-Judge” framework—a new agent evaluation paradigm that’s kind of like getting robots to grade each other’s homework. Where typical evaluation methods focus solely on outcomes or demand extensive manual work, this approach uses agent systems to evaluate agent systems, offering intermediate feedback throughout …

Elder financial exploitation has been a problem for banks for years, and it’s getting worse. Gunther offers practical suggestions for what banks should do when they suspect an older customer is a victim.

We break down OpenAI’s realtime API. Learn how to seamlessly integrate powerful language models into your applications for instant, context-aware responses that drive user engagement. Whether you’re building chatbots, dynamic content tools, or enhancing real-time collaboration, we walk through the API’s capabilities, potential use cases, and best p…

The Arkansas community development financial institution has an ambitious goal of making $500 million worth of mortgages in rural, minority and low-income neighborhoods.

As multi-agent systems grow in importance for fields ranging from customer support to autonomous decision-making, OpenAI has introduced Swarm, an experimental framework that simplifies the process of building and managing these systems. Swarm, a lightweight Python library, is designed for educational purposes, stripping away complex abstractions to…

In this episode, we dive into the intriguing mechanics behind why chat experiences with models like GPT often start slow but then rapidly pick up speed. The key? The KV cache. This essential but under-discussed component enables the seamless and snappy interactions we expect from modern AI systems. Harrison Chu breaks down how the KV cache works, h…

The technology will take on routine, dull work, says Alenka Grealish, principal analyst at Celent.

In this byte-sized podcast, Harrison Chu, Director of Engineering at Arize, breaks down the Shrek Sampler. This innovative Entropy-Based Sampling technique--nicknamed the 'Shrek Sampler--is transforming LLMs. Harrison talks about how this method improves upon traditional sampling strategies by leveraging entropy and varentropy to produce more dynam…

This week, Aman Khan and Harrison Chu explore NotebookLM’s unique features, including its ability to generate realistic-sounding podcast episodes from text (but this podcast is very real!). They dive into some technical underpinnings of the product, specifically the SoundStorm model used for generating high-quality audio, and how it leverages a hie…

Financial institutions have dramatically increased their investment and trust in large language models since 2023. Kartik Ramakrishnan, Capgemini’s deputy CEO of Financial Services and head of banking and capital markets, shares the results of a recent report that analyzed these changes.

OpenAI recently released its o1-preview, which they claim outperforms GPT-4o on a number of benchmarks. These models are designed to think more before answering and handle complex tasks better than their other models, especially science and math questions. We take a closer look at their latest crop of o1 models, and we also highlight some research …

Advances in the company’s AI-based lending models have made them better at predicting risk, which has led to growth, he says.

A recent announcement on X boasted a tuned model with pretty outstanding performance, and claimed these results were achieved through Reflection Tuning. However, people were unable to reproduce the results. We dive into some recent drama in the AI community as a jumping off point for a discussion about Reflection 70B. In 2023, there was a paper wri…

This week, we're excited to be joined by Kyle O'Brien, Applied Scientist at Microsoft, to discuss his most recent paper, Composable Interventions for Language Models. Kyle and his team present a new framework, composable interventions, that allows for the study of multiple interventions applied sequentially to the same language model. The discussio…

Dobrin, founder of advisory firm Qantm AI and former global chief AI officer at IBM, warns that popular generative AI models were trained on the whole of the internet and hallucinate at an unacceptable rate.

This week’s paper presents a comprehensive study of the performance of various LLMs acting as judges. The researchers leverage TriviaQA as a benchmark for assessing objective knowledge reasoning of LLMs and evaluate them alongside human annotations which they find to have a high inter-annotator agreement. The study includes nine judge models and ni…

Jake Donoghue, author of the book Crypto Confidential, shares some of the worst practices he saw as a founder of a cryptocurrency company.

Meta just released Llama 3.1 405B–according to them, it’s “the first openly available model that rivals the top AI models when it comes to state-of-the-art capabilities in general knowledge, steerability, math, tool use, and multilingual translation.” Will the latest Llama herd ignite new applications and modeling paradigms like synthetic data gene…

Karen Petrou, the managing partner at Federal Financial Analytics and a long-time observer of banking and regulation, says banks need to do far more due diligence on potential fintech partners and exert more control over these relationships.

Chaining language model (LM) calls as composable modules is fueling a new way of programming, but ensuring LMs adhere to important constraints requires heuristic “prompt engineering.” The paper this week introduces LM Assertions, a programming construct for expressing computational constraints that LMs should satisfy. The researchers integrated the…

Two veterans and executives at Armed Forces Bank – Tom McLean and Jodi Vickery – share the challenges they see their customers face and new products the bank has rolled out this year to better serve them.

Gene Scalia, the banking lobby’s lawyer on retainer for a potential challenge to Washington’s capital reform effort, discusses the state of administrative law after the overturning of a key legal precedent.

Where adapting LLMs to specialized domains is essential (e.g., recent news, enterprise private documents), we discuss a paper that asks how we adapt pre-trained LLMs for RAG in specialized domains. SallyAnn DeLucia is joined by Sai Kolasani, researcher at UC Berkeley’s RISE Lab (and Arize AI Intern), to talk about his work on RAFT: Adapting Languag…

The Florida bank has emerged from de novo status and can now offer its solar loans in a larger geographic footprint.

It’s been an exciting couple weeks for GenAI! Join us as we discuss the latest research from OpenAI and Anthropic. We’re excited to chat about this significant step forward in understanding how LLMs work and the implications it has for deeper understanding of the neural activity of language models. We take a closer look at some recent research from…

Foundational models like GPT-4, the large language model behind ChatGPT, have hoovered up content from publications like The New York Times and social media sites like Reddit and OpenAI, and it faces several lawsuits because of this. John Thompson, global head of artificial intelligence at EY and author of the book Data for All, has set up what is …

We break down the paper--Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models' Alignment. Ensuring alignment (aka: making models behave in accordance with human intentions) has become a critical task before deploying LLMs in real-world applications. However, a major challenge faced by practitioners is the lack of clear guid…

Proof of identity is critical for many things, including being able to open a bank account, get a job, or obtain health care. Yet proving one’s identity is getting harder in a world of frequent data breaches. We asked Mariana Dahan, founder of the World Identity Network and chair of the Universal ID Council, what she thinks will solve this problem.…

Due to the cumbersome nature of human evaluation and limitations of code-based evaluation, Large Language Models (LLMs) are increasingly being used to assist humans in evaluating LLM outputs. Yet LLM-generated evaluators often inherit the problems of the LLMs they evaluate, requiring further human validation. This week’s paper explores EvalGen, a m…

Custodia Founder and CEO Caitlin Long says the Federal Reserve has rewritten the rules around accessing the government's payments system. The central bank and a federal court judge disagree. Editor’s note: This conversation was recorded on April 17. On April 26, Custodia Bank filed a notice of appeal, signaling that it will challenge the district c…

This week we explore ReAct, an approach that enhances the reasoning and decision-making capabilities of LLMs by combining step-by-step reasoning with the ability to take actions and gather information from external sources in a unified framework. Learn more about AI observability and evaluation, join the Arize AI Slack community or get the latest o…