© 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. 25108
    Prev
    Next
    INDEX
    Explanations

    instances of the word "working."

    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
    426
    +0.11
    0.6%
    429
    +0.11
    0.6%
    406
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    491
    +0.11
    0.04
    246
    +0.11
    0.04
    393
    +0.11
    0.04
    Negative Logits
    lin
    -1.81
    here
    -1.69
    Ī
    -1.65
    cause
    -1.50
    inkle
    -1.50
    ³
    -1.49
    sell
    -1.48
    enth
    -1.47
    azz
    -1.43
    nor
    -1.41
    POSITIVE LOGITS
     grac
    1.63
    igos
    1.51
    cav
    1.51
     comprehension
    1.37
    ======
    1.31
    inux
    1.30
    formaldehyde
    1.29
     mixture
    1.28
    {}{
    1.25
    ongo
    1.24
    Activations Density 0.020%

    No Known Activations