✨ Overview

Shape the future of AI systems through smarter model routing and evaluation! Join us for a groundbreaking hackathon focused on creating judges that evaluate model outputs and routers that direct queries to the optimal models. Whether you're a researcher, developer, or ML enthusiast, this hackathon offers a unique opportunity to tackle crucial challenges in AI optimization and safety.

Judges and routers play a crucial role in AI systems by optimizing decision-making and task delegation. Judges evaluate how well different models perform on specific tasks—not just for correctness, but also across dimensions users increasingly care about, such as ethics, legality, truthfulness, and latency. Routers then direct each query to the most suitable model, aggregating the strengths of many to consistently outperform any single system. This architecture delivers higher-quality, lower-cost responses, while also incentivizing deeper model understanding—since better interpretability leads to better routing.

We’re excited to collaborate with Martian, pioneers in model routing and interpretability research, for this specialized hackathon. Their novel "model mapping" approach distills complex AI models into compact, predictive components that retain only the information needed to estimate performance on specific tasks. By combining this with cutting-edge Mechanistic Interpretability techniques, Martian is developing a general theory of model capabilities—enabling us to evaluate and deploy models more safely, transparently, and effectively. Their work also supports an open ecosystem of specialized models, helping democratize AI development while aligning business impact with AI safety.

Each participating team will receive $50 in model API credits to power their projects, with additional credits available for promising implementations. You'll have access to Martian's judge and router APIs, along with sample code libraries to kickstart your projects.

Sign up here to stay updated for this event

Why This Hackathon Matters:

The Model Mapping Hackathon addresses fundamental challenges in AI deployment as models continue to proliferate:

• Efficiency & Cost Reduction: By routing queries to the most appropriate models, we can dramatically reduce computational costs while improving results

• Democratizing AI Creation: Specialized models excel in narrow domains but struggle with universal tasks - router technology enables these specialized models to thrive

• Safety & Reliability: Better evaluation mechanisms (judges) help ensure outputs meet safety, ethical, and quality standards before deployment

• Understanding Model Capabilities: Developing judges requires deeper mechanistic understanding of how models work and their inherent limitations

Challenge Tracks

Creating Judges Track

1. Model Characteristic Analysis: Create dataset(s) that test an interesting model characteristic pertinent to safety (e.g., ethics, hallucinations, gender bias). Build a judge using this data and evaluate multiple models.

2. Judge Evaluation Metrics: Develop methods to measure judge accuracy, completeness, and reliability for specific characteristics. Explore how this impacts AI Safety.

3. Mechanistic Interpretability for Judges: Apply MI techniques to model internals to create better or more interpretable judges e.g. judges that can evaluate outputs based on how they were generated.

4. Measuring model similarity: HuggingFace hosts thousands of LLMs. How do we measure whether two models have similar capabilities? How do we choose a subset of these models with a sufficiently diverse set of capabilities, that after training the resulting router will be performant? How does the router performance vary with the size of the subset?

Creating Routers Track

Given a set of models with known capabilities measured by known judge scores:

1. Risk-Sensitive Routing: Build efficient routing algorithms balancing the judge scores, computes costs, and system reliability for the best user experience.

2. Multi-Objective Routing: Create routers that use scores from multiple judges (e.g., answer correctness, ethics and legality) according to user preferences for the best user experience. What are the tradeoffs?

3. Routing algorithms: For expensive models, the judge provides a “pre-hoc” way to estimate prediction success (without querying the model). For cheap models, we can ask the model to evaluate the answer, and evaluate its confidence (“post-hoc”) in its predictions. Find interesting ways to mix pre- and post-hoc routing to get the best of both worlds.

4. Multi-level routing: Investigate using a tree of choices rather than one-off routing. What are the pros and cons?

5. Reducing router training costs: Given a model and a task, how can we cheaply detect a model is not a good fit for a task - avoiding further training time optimizing how bad a fit it is.

6. Task Decomposition: Model breaking a complex user task into multiple subtasks that can be routed to the most capable models before recombining the results. What are the AI Safety, cyber-security and/or cost implications of this approach?

7. Universal router: For a set of tasks, create a single router across a set of LLMs that provides higher-quality answers than any single LLM does.

Inferring Judges & Routers Track

1. Reverse Engineering: Given a black-box LLM or router, infer its embedded judge (reward signal) for specific characteristics.

2. Efficiency Analysis: Quantify potential electricity/resource consumption reduction from widespread adoption of optimal routing technologies.

3. Learning when to fail: Sometimes no model will successfully answer a user query. Can we detect when we should fail cheaply?

4. Learning with uncertain signals: How does judge noise affect the router training process? How does noisy feedback data affect the judge/router training process? Is off-policy data a problem when it comes to training routers?

5. Risk sensitivity: Rather than optimizing for expected cost/quality, can we optimize for some other risk profile? E.g. we might tolerate a slightly higher cost and lower quality, if we reduce the variance or minimize a long tail.

6. Create a distilled predictor: The Language Models (Mostly) Know What They Know paper shows that a model can sometimes predict whether it will be able to answer a question correctly. For a selected open “base” model, create a smaller “distilled predictor” that mirrors the base model’s ability to predict answer correctness (but can no longer calculate the answer). You might use the techniques from that paper and/or the pruning and distillation techniques from the movement pruning to shrink the predictor.