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

    Neuronpedia

    Jacobian LensNEW
    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. 10268
    Prev
    Next
    INDEX
    Explanations

    XML or HTML closing tags and document structure elements

    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
    478
    +0.13
    0.7%
    53
    +0.10
    0.6%
    317
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    0
    +0.13
    0.03
    124
    +0.10
    0.03
    374
    +0.10
    0.03
    Negative Logits
    .[]{
    -2.36
    .](
    -2.13
    .]{}
    -1.84
    ).](
    -1.72
    .’
    -1.72
    []{
    -1.71
    .’”
    -1.70
    obacter
    -1.70
    amycin
    -1.69
    *]{}.
    -1.69
    POSITIVE LOGITS
    į
    1.73
    ·
    1.54
     cleft
    1.43
    =>
    1.42
    thead
    1.37
    ĭ
    1.36
    lain
    1.34
    ></
    1.33
     broad
    1.33
     sod
    1.32
    Activations Density 0.088%

    No Known Activations