PERPETUAL
Blog

Rethinking Continual Learning: Learn the Pattern, Not the Correction

July 20264 min read

When an agent gets corrected, the question is not just how to update it. It is what the correction teaches us about the next hundred cases.

What Should AI Agents Learn From?

Our previous post showed that AI agents can improve after long-horizon work. That is only half the problem. Once a trajectory produces an outcome, a review, or an error, the system still has to decide what becomes durable know-how.

It is tempting to say: learn from everything. In production, that breaks down. Good teams do not turn every review note into policy. They ask whether a note is a one-off, a noisy opinion, or a sign of something deeper. Perpetual treats continual learning the same way: find the reusable failure pattern inside noisy work, validate it, and update the right layer of the agent.

Why the Signal Is Hard

In long-horizon work, supervision is often indirect. A reviewer may comment on the whole run, point to one bad step, approve a decision, or simply reveal that the outcome failed. The real world has no evals, only outcomes. The basic loop captures this after-the-fact signal and tries to learn from it.

A simpler agentic learning loop trains directly on that signal. This is useful, but it treats the correction as the lesson. The harder question is whether the same weakness appears elsewhere.

Why Every Correction Is Not Enough

Imagine an employee reviewing an agent that ranks refinance opportunities. The review says:

"This borrower locked a very low fixed rate during the COVID-era rate cycle. Home equity alone is not enough to infer refinance intent."

There is probably something useful in that review. But one review is not enough to change an agent. Before anything becomes durable, three questions matter:

From Correction to Auto-Discover

So the question changes. Instead of asking

How do we learn from this correction?

we should first ask

What hypothesis does this signal support?

That question matters because learning is compression. People do not keep a raw tape of every mistake. The brain filters noise, consolidates facts, turns repeated actions into habits, and uses error signals to ask why an action failed. We do not remember every correction as a separate rule. We learn the pattern that would have prevented the mistake.

Auto-Discover gives agents that missing pause before they learn. It looks at a trajectory and asks: is this a fact, a behavior, a failure mode, or just noise? Promising explanations become candidate hypotheses. Auto-Discover checks them against related examples, drops the ones that do not hold up, and sends the validated pattern to the right place: memory, the agent harness, or weights.

The contrast is simple. A direct loop captures feedback and updates immediately:

Capture feedback and direct update loop A two-node loop cycles between capturing feedback and making a direct update. Capture is gray structure and the update is blue. CAPTURE Feedback LEARN Update
The basic loop trains on the captured signal, but the correction itself may not be the right unit of learning.

Perpetual inserts Auto-Discover before the update:

Auto-Discover test-time search for validated patterns Actions, outcomes, and feedback enter Auto-Discover. Candidate failure hypotheses are simulated, discarded, or validated, and validated patterns exit as fact, behavior, failure, or noise. ACTIONS OUTCOMES FEEDBACK AUTO-DISCOVER LEARNS FROM VALIDATED PATTERNS SIMULATE · DISCARD VALIDATED FACT BEHAVIOR FAILURE NOISE
Auto-Discover searches over candidate failure hypotheses, rejects weak explanations, and keeps only patterns that generalize.

That routing decision is the heart of the system: what to keep, what to ignore, and which layer of the agent should change.

Validated Patterns as the Unit of Learning

After enough work, the signal starts to separate. Across many trajectories, Auto-Discover can distinguish reusable facts, persistent behaviors, repairable failures, and noise. It surfaces patterns such as:

Learning the shape of failure instead of private data Private data examples stay on the customer side of a boundary. Auto-Discover validates transferable facts, behaviors, failures, and noise on the other side. PRIVATE DATA APPLY DISCOUNT ANALYST EDIT TOOL ERROR CUSTOMER BOUNDARY FACT · VALIDATED BEHAVIOR · VALIDATED FAILURE · VALIDATED NOISE · VALIDATED PATTERNS TRANSFER. DATA DOESN'T.
Perpetual learns the shape of failure without turning customer-private data into shared training data.

Once Auto-Discover validates a pattern, it becomes a reusable unit of improvement. One corrected trajectory can seed related tasks, targeted hindsight feedback, and evaluations that check whether the weakness is actually fixed. The result might be a memory entry, a harness change, or a weight update. The key is that the system learns a generalizable pattern, not a private example.

The Perpetual Loop

Together, the two posts describe Perpetual's loop: learn from real work, use Auto-Discover to test what generalizes, and keep only what should become durable know-how. The application is now enterprise agents, but the principle comes from our automated red-teaming work: robust systems improve when they find the mechanism behind a failure. Patterns transfer. Data doesn't.

Today’s AI agents perform work. Perpetual agents learn from the work, and Auto-Discover helps them keep what matters.