When we talk about aligning a Large Language Model (LLM), we're moving beyond traditional metrics like perplexity or accuracy on specific NLP benchmarks. Alignment concerns the extent to which an LLM's behavior consistently matches the intentions of its human designers and users across a diverse range of situations. It's about ensuring the model does what we want it to do, adheres to specified ethical guidelines, and avoids unintended negative consequences.

Core Components of Alignment

While a single, universally accepted definition remains elusive due to the complexity of "intent," alignment is commonly understood through a set of desired behavioral properties. Often, these are summarized as:

Helpfulness: The model should understand and follow user instructions accurately and effectively. It should aim to complete the tasks assigned to it in a useful manner.
Honesty: The model should provide factually accurate information based on its training data. Significantly, it should also represent its own knowledge limitations accurately, avoiding "hallucinations" or confidently stating falsehoods. If uncertain, it should ideally indicate that uncertainty.
Harmlessness: The model must avoid generating outputs that are toxic, discriminatory, promote illegal activities, or cause other forms of harm. This includes adhering to safety protocols and content restrictions defined by its developers.

These three components (Helpful, Honest, Harmless or "HHH") provide a practical framework for evaluating and steering LLM behavior, although the precise interpretation and weighting of each can vary depending on the application context.

A conceptual view of alignment as the congruence between human intentions and observed model behavior, often operationalized through properties like helpfulness, honesty, and harmlessness.

Alignment vs. Capability

It's important to distinguish alignment from the model's raw capabilities. Capabilities refer to the model's inherent abilities learned during pre-training, such as understanding grammar, storing factual knowledge, reasoning, or generating creative text. A model can be highly capable but poorly aligned (e.g., skillfully generating harmful content when prompted) or less capable but reasonably aligned (e.g., refusing harmful requests but also struggling with complex instructions). Alignment techniques, which we will examine throughout this course, aim to steer the model's existing capabilities towards desired outcomes.

Towards a Formal View

Conceptually, we can think of alignment as minimizing the divergence between the probability distribution of the model's outputs, $P_{\text{model}}(y | x)$ , and an idealized distribution representing the intended behavior, $P_{\text{intended}}(y | x)$ , given some input context $x$ .

\text{Minimize } D(P_{\text{model}}(y | x) || P_{\text{intended}}(y | x))

Here, $D$ represents some measure of divergence (like KL divergence). The significant challenge, as discussed in the next section, is that $P_{\text{intended}}$ is extremely difficult, if not impossible, to specify completely and accurately for all possible inputs $x$ . It embodies complex human preferences, ethical norms, and contextual nuances. Much of the work in alignment involves finding effective proxies and methods to approximate and optimize towards this intended distribution.

Whose Intent?

Furthermore, "human intent" isn't monolithic. Developers might prioritize safety and robustness, deployers might focus on specific application goals and brand reputation, while end-users have immediate task objectives. Society at large has expectations regarding fairness, bias, and long-term impacts. Balancing these potentially conflicting intentions is a central part of the alignment challenge.

Understanding this definition of alignment, centered on intended behavior rather than just task performance, sets the stage for appreciating the difficulties involved (the "Alignment Problem") and the necessity of specialized techniques like Reinforcement Learning from Human Feedback (RLHF) and others we will cover.