The Anthropic papers “Towards monosemanticity” and “Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet” demonstrate how sparse autoencoders can extract interpretable features from large language models, converting polysemantic neuron activations into monosemantic representations that directly map to identifiable concepts and behaviors. In this writeup I try to and explain the core concepts in this research.
A sparse autoencoder is a neural network designed to learn a compact, interpretablerepresentation of input data by enforcing sparsity on its hidden layer activations. A sparse autoencoder is “sparse” because it applies a constraint during training so that, for any given input, only a small subset of the hidden (latent) units is active (nonzero). This is achieved by adding a sparsity penalty to the loss function, commonly L1 regularization or a KL-divergence term, which discourages most activations from deviating much from zero. This ensures the encoded representation is sparse—meaning only a few features are used to reconstruct the input—resulting in greater interpretability and the extraction of meaningful features. It is an “autoencoder” because the full model is trained end-to-end to reconstruct its own input. The encoder maps the input data to a latent code, and the decoder maps it back to the reconstruction. The central training objective is to minimize reconstruction error, making the network learn to reproduce its input as closely as possible. The difference from other autoencoder types (e.g., vanilla, denoising, variational) is specifically the addition of the sparsity constraint on the hidden code.
An activation is the output value of a neuron or unit in a neural network layer after applying an activation function to a weighted sum of inputs. Mathematically, for a neuron receiving inputs x1,x2,…,xnx1,x2,…,xn with weights w1,w2,…,wnw1,w2,…,wn, the activation is a=f(w1x1+w2x2+⋯+wnxn+b)a=f(w1x1+w2x2+⋯+wnxn+b), where ff is the activation function (such as ReLU, sigmoid, or tanh) and bb is a bias term.
The idea is to view activations as superpositions of underlying features and to use a neural network to reverse the mapping from the activations to the features. This is peering into the workings of an LLM with another neural network to see what the activations mean.
So in the monosemanticity quest, the activations are seen as a superposition of underlying features. A sparse autoencoder decomposes model activations into interpretable features by expressing each activation vector as a sparse linear combination of learned feature directions. Given an activation vector xjxj, the decomposition is:xj≈b+∑ifi(xj)dixj≈b+i∑fi(xj)di where fi(xj)fi(xj) is the activation (magnitude) of feature ii, didi is a unit vector representing the direction of feature ii in activation space, and bb is a bias term. The feature activations are computed by the encoder as fi(x)=ReLU(We(x−bd)+be)ifi(x)=ReLU(We(x−bd)+be)i, where WeWe is the encoder weight matrix and bdbd, bebe are pre-encoder and encoder biases. The feature directions are the columns of the decoder weight matrix WdWd. This formulation is dictionary learning: each activation is reconstructed from a sparse set of learned basis vectors scaled by their respective feature activations.
Acts is short for activations in the above figure of a sparse auto encoder functioning from Anthropic. .
Does the SAE look at all the activations or only certain layers ?
Sparse autoencoders are typically trained on activations from specific layers rather than all layers simultaneously. In practice, a separate SAE is trained for each layer or location in the model where one wishes to analyze or intervene on activations. In Anthropic’s “Scaling Monosemanticity” paper specifically, the SAE was trained only on activations from the residual stream at the middle layer (halfway through Claude 3 Sonnet). This choice was made for several reasons: the residual stream is smaller than the MLP layer, making training and inference computationally cheaper; focusing on the residual stream mitigates “cross-layer superposition,” which refers to neurons whose activations depend on combinations of information across multiple layers; and the middle layer likely contains more interesting and abstract features compared to early layers (which capture basic patterns) or final layers (which may be too task-specific).
Motivation and Definitions
- Large language models (LLMs) typically exhibit polysemantic neurons, which activate in response to numerous, often unrelated, concepts, impeding interpretability and safe control.
- Monosemanticity refers to representations where each learned feature corresponds to a single, easily identifiable concept, thus improving transparency in model operations.
- Sparse autoencoders (SAEs) are employed to learn dictionary-like decompositions of hidden activations, aiming for each basis vector (feature) to align with a distinct semantic unit rather than mixed signals.
Methods and Techniques
- The approach uses SAEs to project model activations into higher-dimensional, sparse spaces where individual features become interpretable.
- Dictionary learning is central: activations from a given layer are encoded by the SAE so that each dictionary element ideally corresponds to a unique concept or pattern.
- Anthropic scales this method from small, shallow models to large networks by training SAEs on billions of activations from state-of-the-art LLMs (e.g., Claude 3 Sonnet).
- Modifying feature coefficients within the SAE’s learned space causes proportional, causal shifts in the model’s reconstructed activation, allowing direct steering of outputs at runtime.
- Feature steering leverages these interpretable directions to alter specific model behaviors (e.g., changing model goals, tone, biases, or inducing controlled errors) by adjusting activation values during inference.
Results and Empirical Findings
- The method yields dictionaries where a substantial portion of features (by human evaluation, approximately 70%) are monosemantic—associated with singular, nameable concepts such as DNA motifs or language script.
- Quantitative validation includes human raters agreeing with feature names, decoder-row alignment (cosine similarity > 0.86 between encoder and decoder vectors), and strong compositionality in steering outcomes.
- Scaling up the size of the SAE dictionary increases the proportion of monosemantic features and the precision of behavioral interventions.
- Interventions using these features show robust control over model outputs, evidenced by targeted behavioral scores and ability to suppress or augment specific behaviors with tunable steering coefficients.
Conceptual Advances
- The work empirically supports the superposition hypothesis: raw neurons entangle multiple meanings, but sparse dictionary learning untangles these into separately addressable features.
- The method demonstrates that high-dimensional, sparsely coded representations can be extracted at scale without significant algorithmic changes, opening new paths for mechanistic interpretability and control tools in LLMs.
- These advances suggest dictionary learning could, in future, replace large fine-tuning campaigns for behavioral adjustments, increase safety monitoring, and allow new forms of user-customized steering.
Activation Steering and Implications
- Steering methods operate by selecting, amplifying, or suppressing identified sparse features using signed, tunable coefficients (λλ), with each adjustment reflected directly and causally in output behavior.
- The process is mathematically tractable because the SAE remains linear; interventions can be analyzed for causal effects and compositional interactions, which is not feasible in the dense activation spaces of standard LLMs.
- This enables multifaceted interventions and targeted control: steering vectors can increase or decrease model propensities for specific behaviors, factuality, style, or compliance in a transparent manner.
Summary Table: Key Terms
This research establishes scalable techniques for extracting and manipulating interpretable features in large LLMs, enabling precise behavioral steering and laying groundwork for safer, more controllable AI deployments.
The sparse autoencoder (SAE) in Anthropic’s “Scaling Monosemanticity” paper was trained at three different scales on activations from Claude 3 Sonnet: approximately 1 million (1,048,576), 4 million (4,194,304), and 34 million (33,554,432) features. For the largest run, the 34M-feature SAE, the number of active (nonzero) features for any given token was typically fewer than 300, showing high sparsity.
The paper emphasizes that many extracted features are relevant to AI safety, such as features for security vulnerabilities, code backdoors, bias (overt and subtle), deception (including power-seeking and treacherous turns), sycophancy, and the generation of dangerous or criminal content. However, the authors note that the detection of such features is preliminary and should not be over-interpreted: knowing about harmful behaviors is distinct from enacting them. The presence of potentially dangerous features suggests the model could represent these concepts internally, warranting deeper investigation. The interpretability gained through the SAE allows for the identification and possible intervention on such features but does not automatically ensure safe model behavior without further work and robust evaluation.
The authors compare their feature-extraction approach to previous interpretability and model-steering methods:
- Unlike neuron-centric methods, which often yield tangled, polysemantic activations, SAEs learn overcomplete, sparse dictionaries that approximate monosemantic features.
- Their approach leverages scaling laws to optimize both the number of features and training steps, showing that larger SAEs provide more granular, precise, and interpretable decompositions than smaller or denser models.
- The SAE-based approach allows for explicit, steerable interventions by clamping or zeroing specific features, something not possible with conventional dense neuron manipulation.
- The paper positions this technique as extensible, mechanistically transparent, and a foundation for scalable model interpretability—offering capabilities not matched by most prior strategies.
These results highlight that scalable, sparse autoencoders produce directly actionable, interpretable features offering new tools for AI safety and more precise model control compared to traditional neuron or layerwise interpretability approaches.
An argument on the urgency of interpretability: https://www.darioamodei.com/post/the-urgency-of-interpretability
Neel Nanda’s replication of results has a notebook for going deeper. https://www.alignmentforum.org/posts/fKuugaxt2XLTkASkk/open-source-replication-and-commentary-on-anthropic-s
Secure Machinery 