© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. MLP Post
    9. 0-MLP-SM
    11. 30997
    Prev
    Next
    INDEX
    Explanations

    terms related to specialized fields or expertise

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_mlp_out
    Hook Layer
    0
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.35
    2.1%
    148
    +0.14
    0.9%
    29
    +0.14
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    29
    +0.35
    0.01
    380
    +0.14
    0.01
    156
    +0.14
    0.01
    Negative Logits
    enson
    -1.51
    raised
    -1.45
    itsu
    -1.44
    rox
    -1.44
     conjecture
    -1.37
    aned
    -1.35
    ros
    -1.33
    hol
    -1.33
    yo
    -1.31
    aled
    -1.23
    POSITIVE LOGITS
    positions
    1.55
    posing
    1.52
    izing
    1.51
    amente
    1.47
     needs
    1.47
    {}{
    1.44
     powers
    1.43
    quarters
    1.43
    ently
    1.39
    istine
    1.34
    Activations Density 0.030%

    No Known Activations