AI Evaluation Methods

BREAKING! The FDA just released this draft guidance, titled Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations, that aims to provide industry and FDA staff with a Total Product Life Cycle (TPLC) approach for developing, validating, and maintaining AI-enabled medical devices. The guidance is important even in its draft stage in providing more detailed, AI-specific instructions on what regulators expect in marketing submissions; and how developers can control AI bias. What’s new in it? 1) It requests clear explanations of how and why AI is used within the device. 2) It requires sponsors to provide adequate instructions, warnings, and limitations so that users understand the model’s outputs and scope (e.g., whether further tests or clinical judgment are needed). 3) Encourages sponsors to follow standard risk-management procedures; and stresses that misunderstanding or incorrect interpretation of the AI’s output is a major risk factor. 4) Recommends analyzing performance across subgroups to detect potential AI bias (e.g., different performance in underrepresented demographics). 5) Recommends robust testing (e.g., sensitivity, specificity, AUC, PPV/NPV) on datasets that match the intended clinical conditions. 6) Recognizes that AI performance may drift (e.g., as clinical practice changes), therefore sponsors are advised to maintain ongoing monitoring, identify performance deterioration, and enact timely mitigations. 7) Discusses AI-specific security threats (e.g., data poisoning, model inversion/stealing, adversarial inputs) and encourages sponsors to adopt threat modeling and testing (fuzz testing, penetration testing). 8) And proposed for public-facing FDA summaries (e.g., 510(k) Summaries, De Novo decision summaries) to foster user trust and better understanding of the model’s capabilities and limits.

Evaluating reasoning models is non-trivial. But you can use a verifier to check if answers are actually correct. I just finished a new 35-page chapter of Build a Reasoning Model (From Scratch), which is all about building such a verifier from the ground up. Symbolic parsing, math equivalence, edge cases… this was quite the project. But it’s now submitted and will hopefully appear soon on Manning’s Early Access platform. This chapter also includes a recap of other popular evaluation methods (multiple-choice, leaderboards, and judges): 3.1 Understanding the main evaluation methods for LLMs 3.1.1 Evaluating answer-choice accuracy 3.1.2 Using verifiers to check answers 3.1.3 Comparing models using preferences and leaderboards 3.1.4 Judging responses with other LLMs 3.2 Building a math verifier 3.3 Loading a pre-trained model to generate text 3.4 Implementing a wrapper for easier text generation 3.5 Extracting the final answer box 3.6 Normalizing the extracted answer 3.7 Verifying mathematical equivalence 3.8 Grading answers 3.9 Loading the evaluation dataset 3.10 Evaluating the model The code and sneak peak are on GitHub: 📖 https://mng.bz/lZ5B 🔗 https://lnkd.in/g8_7WtRX

As we transition from traditional task-based automation to 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀, understanding 𝘩𝘰𝘸 an agent cognitively processes its environment is no longer optional — it's strategic. This diagram distills the mental model that underpins every intelligent agent architecture — from LangGraph and CrewAI to RAG-based systems and autonomous multi-agent orchestration. The Workflow at a Glance 1. 𝗣𝗲𝗿𝗰𝗲𝗽𝘁𝗶𝗼𝗻 – The agent observes its environment using sensors or inputs (text, APIs, context, tools). 2. 𝗕𝗿𝗮𝗶𝗻 (𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 𝗘𝗻𝗴𝗶𝗻𝗲) – It processes observations via a core LLM, enhanced with memory, planning, and retrieval components. 3. 𝗔𝗰𝘁𝗶𝗼𝗻 – It executes a task, invokes a tool, or responds — influencing the environment. 4. 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 (Implicit or Explicit) – Feedback is integrated to improve future decisions.     This feedback loop mirrors principles from: • The 𝗢𝗢𝗗𝗔 𝗹𝗼𝗼𝗽 (Observe–Orient–Decide–Act) • 𝗖𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲𝘀 used in robotics and AI • 𝗚𝗼𝗮𝗹-𝗰𝗼𝗻𝗱𝗶𝘁𝗶𝗼𝗻𝗲𝗱 𝗿𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 in agent frameworks Most AI applications today are still “reactive.” But agentic AI — autonomous systems that operate continuously and adaptively — requires: • A 𝗰𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 𝗹𝗼𝗼𝗽 for decision-making • Persistent 𝗺𝗲𝗺𝗼𝗿𝘆 and contextual awareness • Tool-use and reasoning across multiple steps • 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 for dynamic goal completion • The ability to 𝗹𝗲𝗮𝗿𝗻 from experience and feedback    This model helps developers, researchers, and architects 𝗿𝗲𝗮𝘀𝗼𝗻 𝗰𝗹𝗲𝗮𝗿𝗹𝘆 𝗮𝗯𝗼𝘂𝘁 𝘄𝗵𝗲𝗿𝗲 𝘁𝗼 𝗲𝗺𝗯𝗲𝗱 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 — and where things tend to break. Whether you’re building agentic workflows, orchestrating LLM-powered systems, or designing AI-native applications — I hope this framework adds value to your thinking. Let’s elevate the conversation around how AI systems 𝘳𝘦𝘢𝘴𝘰𝘯. Curious to hear how you're modeling cognition in your systems.

The NYT just reported that patients are uploading entire medical records into chatbots - but the risks are not what most people think. Patients are pasting labs, imaging, clinical notes, and oncology reports directly into LLMs. • 26-year-old told her labs “most likely” indicated a pituitary tumor. MRI: normal • 63-year-old advised to escalate to catheterization. Found ~85% LAD stenosis Because of how the chatbot responds, many assume the AI reasons about their symptoms and medical record the same way a clinician does. But AI systems are capable of both meaningful help and serious error, without any calibration signal visible to the user. Most worry about wrong AI recommendations. But the bigger risk is what the AI does not say. 📊 Harm preprint study A new Stanford-Harvard study (David Wu, MD, PhD, Fateme (Fatima) Nateghi, Adam Rodman, Jonathan H. Chen et al.) evaluated 31 models on 100 real outpatient eConsult cases across 10 specialties: - 4,249 management actions - 12,747 expert ratings Severe harm per 100 cases: - Best models: ~12–15 - Worst models: ~40 ~77% of severe harms were omissions: - Not ordering a critical test - Missing a needed referral - Neglecting follow-up suggestions 🔷 Additional findings: 1) Top models outperformed generalists using conventional resources (though these were difficult eConsult cases that PCPs were posing to specialists) 2) No link between safety and model size, recency, “reasoning modes,” or standard benchmarks 3) Multi-agent + RAG approaches reduced harm; heterogeneous ensembles had ~6× higher odds of top-quartile safety 📌 Implications When a patient asks AI for medical advice, the primary risk is not incorrect recommendations. It's neglecting critical actions a clinician might suggest (notably, humans also make a lot of mistakes). ⚠️ Why this matters 1) 2/3 of US physicians report using LLMs, and millions of patients. Errors will become more subtle as models get better. Both harms of omissions and commission will become harder for clinicians (and especially patients) to detect. 2) Sampling a few outputs is not enough: clinical AI evaluation needs explicit, systematic harm measurement on real cases, not just performance or accuracy on knowledge benchmarks. 3) If we don’t measure omission harms, we will systematically underestimate risk. 🔴 Open Call: State of Clinical AI Report (Jan 2026) The ARISE Network (Stanford + Harvard) is compiling a State of Clinical AI Report for 2026. Audience: health system leaders, clinicians, researchers, tech/pharma, media, investors 2025 peer reviewed and preprint studies within scope: • Clinical AI (doctor- or patient-facing) • Benchmarks, evaluations, real-world deployments, prospective trials • Workflow, outcomes, and implementation studies 📅 Submission deadline: Dec 21, 2025 - Comment with study link + 1–2 sentences on key findings and why it matters - We will follow up with a one-slide reference example for invited submissions

Reliability, evaluation, and “hallucination anxiety” are where most AI programmes quietly stall. Not because the model is weak. Because the system around it is not built to scale trust. When companies move beyond demos, three hard questions appear: →Can we rely on this output? →Do we know what “good” actually looks like? →How much human oversight is enough? The fix is not better prompting. It is a strategy and operating discipline. 𝐅𝐢𝐫𝐬𝐭: ⁣Define reliability like a product, not a vibe. Every serious AI use case should have a one-page SLO sheet with measurable targets across: →Task success ↳Right-first-time rate and rubric-based acceptance →Factual grounding ↳Evidence coverage and unsupported-claim tracking →Safety and compliance ↳Policy violations and PII leakage →Operational quality ↳Latency, cost per task, escalation to humans Now “good” is no longer opinion. It is observable. 𝐒𝐞𝐜𝐨𝐧𝐝:  evaluation must be continuous, not a one-off demo test. Use a simple loop: 𝐏lan: Define rubrics, datasets, and risk tiers 𝐃⁣o: Run offline evaluations and limited pilots 𝐂heck: Monitor drift and regressions weekly 𝐀ct: Update prompts, data, guardrails, and workflows Support this with an AI test pyramid: →Unit checks for prompts and tool behaviour →Scenario tests for real edge failures →Regression benchmarks to prevent backsliding →Live monitoring in production Add statistical control charts, and you can detect silent degradation before users do. 𝐓𝐡𝐢𝐫𝐝: reduce hallucinations by design. →Run a short failure-mode workshop and engineer controls: →Require retrieval or evidence before answering →Allow safe abstention instead of confident guessing →Add claim checking and tool validation →Use structured intake and clarifying flows You are not asking the model to behave. You are designing a system that expects failure and contains it. 𝐅𝐨𝐮𝐫𝐭𝐡: make human-in-the-loop affordable. Tier risk: →Low risk: Light sampling →Medium risk: Triggered review →High risk: Mandatory approval Escalate only when signals demand it: low confidence, missing evidence, policy flags, or novelty spikes. Review becomes targeted, fast, and a source of improvement data. 𝐅𝐢𝐧𝐚𝐥𝐥𝐲: Operate it like a capability. Track outcomes, risk, delivery speed, and cost on a single dashboard. Hold a short weekly reliability stand-up focused on regressions, failure modes, and ownership. What you end up with is simple: ↳Use case catalogue with risk tiers ↳Clear SLOs and error budgets ↳Continuous evaluation harness ↳Built-in controls ↳Targeted human review ↳Reliability cadence AI does not scale on intelligence alone. It scales on measurable trust. ♻️ Share if you found thisuseful. ➕ Follow (Jyothish Nair) for reflections on AI, change, and human-centred AI #AI #AIReliability #TrustAtScale #OperationalExcellence

The new consulting edge isn't AI. It's knowing when your AI is wrong. Every consultant has been there: You ask AI to analyze documents and generate insights. During review, you spot a questionable stat that doesn't exist in the source! AI hallucinations are a problem. The solution? Implementing "prompt evals". → Prompt evals: directions that force AI to verify its own work before responding. A formula for effective evals: 1. Assign a verification role → "Act as a critical fact-checker whose reputation depends on accuracy" 2. Specify what to verify → "Check all revenue projections against the quarterly reports in the appendix" 3. Define success criteria → "Include specific page references for every statistic" 4. Establish clear terminology → "Rate confidence as High/Medium/Low next to each insight" Here is how your prompt will change: OLD: "Analyze these reports and identify opportunities." NEW: "You are a senior analyst known for accuracy. List growth opportunities from the reports. For each insight, match financials to appendix B, match market claims to bibliography sources, add page ref + High/Med/Low confidence, otherwise write REQUIRES VERIFICATION.” Mastering this takes practice, but the results are worth it. What AI leaders know that most don't: "If there is one thing we can teach people, it's that writing evals is probably the most important thing." Mike Krieger, Anthropic CPO By the time most learn basic prompting, leaders will have turned verification into their competitive advantage. Steps to level-up your eval skills: → Log hallucinations in a "failure library" → Create industry-specific eval templates → Test evals with known error examples → Compare verification with competitors Next time you're presented with AI-generated analysis, the most valuable question isn't about the findings themselves, but: 'What evals did you run to verify this?' This simple inquiry will elevate your teams approach to AI & signal that in your organization, accuracy isn't optional.

You've built your AI agent... but how do you know it's not failing silently in production? Building AI agents is only the beginning. If you’re thinking of shipping agents into production without a solid evaluation loop, you’re setting yourself up for silent failures, wasted compute, and eventully broken trust. Here’s how to make your AI agents production-ready with a clear, actionable evaluation framework: 𝟭. 𝗜𝗻𝘀𝘁𝗿𝘂𝗺𝗲𝗻𝘁 𝘁𝗵𝗲 𝗥𝗼𝘂𝘁𝗲𝗿 The router is your agent’s control center. Make sure you’re logging: - Function Selection: Which skill or tool did it choose? Was it the right one for the input? - Parameter Extraction: Did it extract the correct arguments? Were they formatted and passed correctly? ✅ Action: Add logs and traces to every routing decision. Measure correctness on real queries, not just happy paths. 𝟮. 𝗠𝗼𝗻𝗶𝘁𝗼𝗿 𝘁𝗵𝗲 𝗦𝗸𝗶𝗹𝗹𝘀 These are your execution blocks; API calls, RAG pipelines, code snippets, etc. You need to track: - Task Execution: Did the function run successfully? - Output Validity: Was the result accurate, complete, and usable? ✅ Action: Wrap skills with validation checks. Add fallback logic if a skill returns an invalid or incomplete response. 𝟯. 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝘁𝗵𝗲 𝗣𝗮𝘁𝗵 This is where most agents break down in production: taking too many steps or producing inconsistent outcomes. Track: - Step Count: How many hops did it take to get to a result? - Behavior Consistency: Does the agent respond the same way to similar inputs? ✅ Action: Set thresholds for max steps per query. Create dashboards to visualize behavior drift over time. 𝟰. 𝗗𝗲𝗳𝗶𝗻𝗲 𝗦𝘂𝗰𝗰𝗲𝘀𝘀 𝗠𝗲𝘁𝗿𝗶𝗰𝘀 𝗧𝗵𝗮𝘁 𝗠𝗮𝘁𝘁𝗲𝗿 Don’t just measure token count or latency. Tie success to outcomes. Examples: - Was the support ticket resolved? - Did the agent generate correct code? - Was the user satisfied? ✅ Action: Align evaluation metrics with real business KPIs. Share them with product and ops teams. Make it measurable. Make it observable. Make it reliable. That’s how enterprises scale AI agents. Easier said than done.

The hype around AI often leads people to overestimate its capabilities. They copy-paste responses. Accept every output as fact. But here's what 67% of business leaders refuse to acknowledge⤵️ - AI is brilliant at lying with confidence. If you want real AI value, you need two things: powerful tools and human judgment. And critical thinking that questions every output, every source, every "fact." Great AI implementation doesn't just generate content. - It amplifies human expertise. - It speeds up verification. Pattern recognition will show you connections, but wisdom is what separates signal from noise. So, if you're ready for transformation, here's a battle-tested framework to get ahead: - Treat AI like a brilliant intern. - Assume every output needs fact-checking. Create verification protocols before you deploy any AI-generated content or insights. - Build validation workflows. - Cross-reference sources. - Check citations. Verify statistics against original research. Make skepticism your default mode. Layer human expertise. Use AI to accelerate research, not replace thinking. Subject matter experts should review, refine, and approve all critical outputs. 🔻 Create feedback loops. - Track where AI gets it wrong. - Build those learnings back into your prompts and processes. - Failed outputs teach you as much as successful ones. 🔻 Invest in verification tools. - Dedicate 30% of your AI budget to fact-checking systems, source validation, and human oversight. - Prevention costs less than correction. Every output gets both. 👉 Combine algorithmic power with human wisdom. That's how you harness AI without getting burned by its confident hallucinations. Are you ready to double-check AI's work, or will you take it at face value?

𝔼𝕍𝔸𝕃 field note (2 of 3): Finding the benchmarks that matter for your own use cases is one of the biggest contributors to AI success. Let's dive in. AI adoption hinges on two foundational pillars: quality and trust. Like the dual nature of a superhero, quality and trust play distinct but interconnected roles in ensuring the success of AI systems. This duality underscores the importance of rigorous evaluation. Benchmarks, whether automated or human-centric, are the tools that allow us to measure and enhance quality while systematically building trust. By identifying the benchmarks that matter for your specific use case, you can ensure your AI system not only performs at its peak but also inspires confidence in its users. 🦸♂️ Quality is the superpower—think Superman—able to deliver remarkable feats like reasoning and understanding across modalities to deliver innovative capabilities. Evaluating quality involves tools like controllability frameworks to ensure predictable behavior, performance metrics to set clear expectations, and methods like automated benchmarks and human evaluations to measure capabilities. Techniques such as red-teaming further stress-test the system to identify blind spots. 👓 But trust is the alter ego—Clark Kent—the steady, dependable force that puts the superpower into the right place at the right time, and ensures these powers are used wisely and responsibly. Building trust requires measures that ensure systems are helpful (meeting user needs), harmless (avoiding unintended harm), and fair (mitigating bias). Transparency through explainability and robust verification processes further solidifies user confidence by revealing where a system excels—and where it isn’t ready yet. For AI systems, one cannot thrive without the other. A system with exceptional quality but no trust risks indifference or rejection - a collective "shrug" from your users. Conversely, all the trust in the world without quality reduces the potential to deliver real value. To ensure success, prioritize benchmarks that align with your use case, continuously measure both quality and trust, and adapt your evaluation as your system evolves. You can get started today: map use case requirements to benchmark types, identify critical metrics (accuracy, latency, bias), set minimum performance thresholds (aka: exit criteria), and choose complementary benchmarks (for better coverage of failure modes, and to avoid over-fitting to a single number). By doing so, you can build AI systems that not only perform but also earn the trust of their users—unlocking long-term value.

We have to internalize the probabilistic nature of AI. There’s always a confidence threshold somewhere under the hood for every generated answer and it's important to know that AI doesn’t always have reasonable answers. In fact, occasional "off-the-rails" moments are part of the process. If you're an AI PM Builder (as per my 3 AI PM types framework from last week) - my advice: 1. Design for Uncertainty: ✨Human-in-the-loop systems: Incorporate human oversight and intervention where necessary, especially for critical decisions or sensitive tasks. ✨Error handling: Implement robust error handling mechanisms and fallback strategies to gracefully manage AI failures (and keep users happy). ✨User feedback: Provide users with clear feedback on the confidence level of AI outputs and allow them to provide feedback on errors or unexpected results. 2. Embrace an experimental culture & Iteration / Learning: ✨Continuous monitoring: Track the AI system's performance over time, identify areas for improvement, and retrain models as needed. ✨A/B testing: Experiment with different AI models and approaches to optimize accuracy and reliability. ✨Feedback loops: Encourage feedback from users and stakeholders to continuously refine the AI product and address its limitations. 3. Set Realistic Expectations: ✨Educate users: Clearly communicate the potential for AI errors and the inherent uncertainty involved about accuracy and reliability i.e. you may experience hallucinations.. ✨Transparency: Be upfront about the limitations of the system and even better, the confidence levels associated with its outputs.