ai
A biweekly podcast where hosts Nathan Labenz and Erik Torenberg interview the builders on the edge of AI and explore the dramatic shift it will unlock in the coming years. The Cognitive Revolution is part of the Turpentine podcast network. To learn more: turpentine.co
Anthropic has released a new paper detailing "J-Space" and the "J-lens," a method for probing how large language models store and process concepts. The J-lens can identify directions in a model's latent space that correlate with specific token outputs, offering a potential window into the model's internal reasoning processes. This new approach shows promise in understanding complex LLM behaviors.
Attempts to manipulate LLM behavior using the J-lens have shown varied success, with interventions only reliably influencing outcomes 50-70% of the time. The remaining 30-45% of cases result in unpredictable or nonsensical changes, indicating that aspects of LLM cognition remain opaque. Researchers suggest this suggests 'dark matter or dark cognition' within these models.
A novel training method called 'counterfactual reflection' has been introduced, where AI models are paused mid-task and trained to respond according to desired values. This method involves interrupting the model and guiding it toward a 'constitutionally right' response, which is then used for supervised training. Anthropic's research suggests this technique helps embed desired concepts into the J-space, improving overall model behavior.
The J-space probing technique is noted for its low computational cost, making it potentially suitable for production environments. The process involves simple matrix multiplication on layer activations, with an estimated compute overhead of 5% or less, similar to other monitoring methods like constitutional classifiers.
Liquid AI, a company spun out of MIT, is developing device-native foundation models designed for edge devices with limited processing power. CEO Ramin Hassani highlighted that the $800 billion global smartphone and laptop market represents a massive opportunity, especially given concerns about privacy and data control. The company's approach focuses on efficient, biologically inspired algorithms.
Liquid AI has demonstrated the capability of running a 1 billion parameter model on an iPhone, showcasing its "liquid philosophy" for efficient AI. CEO Ramin Hassani mentioned that this model, featuring attention layers and a simple gated learned convolution, can operate fast enough for basic use cases like private document classification.
Liquid AI is refining its network architecture search process by evaluating models on actual hardware and downstream tasks, moving away from potentially misleading proxy metrics. Ramin Hassani explained that for specific use cases and limited compute resources, their search is more likely to discover novel, exotic architectures.
Liquid AI is planning to launch a new platform that will allow customers to fine-tune small AI models for their specific use cases on a self-serve basis. Host anticipates this will ease demand for frontier models and improve global AI access, expressing excitement for its upcoming release.
Liquid AI's core mission, originating from MIT research, is to maximize the intelligence that can be encoded into the smallest algorithmic formats, with efficiency as its cornerstone. Ramin Hassani elaborated that their work on "Liquid neural networks" aims to deliver the reliability of larger AI systems while running on smaller, embedded hardware like CPUs and NPUs.
Liquid AI is developing alternative machine learning algorithms inspired by brain dynamics to achieve better out-of-distribution generalization, a key challenge in real-world applications like robotics. Ramin Hassani explained their research into "Liquid neural networks" aims to mimic how neurons exchange information, potentially unlocking greater capabilities than traditional artificial neural networks.
Liquid AI's foundational research into neural networks was inspired by studying the nervous system of a specific worm, which, with only 300 cells, exhibits complex sensory reactive behavior. Ramin Hassani noted that this biological model offered a more manageable starting point for applying learning theory to systems with far fewer neurons than traditional artificial networks.
Liquid AI has achieved significant market validation, holding the fifth position on the Hugging Face US downloads leaderboard and securing notable partnerships with Shopify and Mercedes-Benz. These achievements underscore the company's progress in commercializing its efficient AI technology.
Thomas Von Schamer, co-founder of Neural Concept, discussed how AI is revolutionizing automotive design. Traditional methods involving physical prototypes and lengthy computer simulations are being replaced by AI models that can provide results in minutes, allowing for the testing of thousands of designs per day. This accelerates product cycles and enables greater innovation.
Thomas Von Schamer of Neural Concept explained that the AI-driven acceleration of engineering processes is a pattern seen across various domains, similar to protein folding. Neural Concept is developing specialist models for physics-based simulations like aerodynamics, crash safety, and thermal management, aiming for general-purpose foundation models for engineering.
Neural Concept's AI models enable manufacturers like Jaguar Land Rover to perform aerodynamic testing on over 1,000 designs per day. This contrasts with traditional methods that were limited to far fewer iterations due to the lengthy simulation times. The AI approach significantly boosts efficiency and allows engineers to explore a wider design space.
Thomas Von Schamer anticipates that the integration of AI into engineering will lead to faster product cycles and an explosion of new form factors with improved quality and efficiency. He likens the pattern of AI development to protein folding, where complex problems are solved more rapidly through data and learning.
Mercury has introduced a new conversational interface called 'Command,' designed to provide natural language access to financial data and enable AI agents to perform actions. This aims to overcome the limitations of traditional banking interfaces for AI integration, offering a more seamless and secure way for businesses to manage their finances.