© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 28513
    Prev
    Next
    INDEX
    Explanations

    references to footnotes and annotations in documents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    442
    +0.17
    1.0%
    198
    +0.15
    0.8%
    17
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    442
    +0.17
    0.13
    7
    +0.15
    0.00
    475
    +0.14
    0.05
    Negative Logits
    ²
    -2.95
    µ
    -2.94
    §
    -2.87
    Ń
    -2.78
    ©
    -2.72
    ¯
    -2.67
      ↵
    -2.65
    ↵
    -2.65
    <|outofrange|>
    -2.65
    č↵
    -2.65
    POSITIVE LOGITS
     Beau
    1.55
     Universal
    1.49
    )?
    1.48
     Ocean
    1.47
     Guild
    1.44
     Parish
    1.42
     Broadcasting
    1.41
     Craft
    1.40
     Medium
    1.40
     Ski
    1.40
    Activations Density 3.720%

    No Known Activations