© 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. 2-MLP-SM
    11. 22477
    Prev
    Next
    INDEX
    Explanations

    proper names and references to individuals, particularly in the context of legal and political situations

    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/2-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.2.hook_mlp_out
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    23
    +0.19
    1.1%
    369
    +0.12
    0.7%
    400
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.19
    0.07
    332
    +0.12
    0.09
    400
    +0.12
    0.06
    Negative Logits
    ater
    -1.46
    naire
    -1.43
     programme
    -1.38
    isme
    -1.37
    guard
    -1.32
     documentary
    -1.32
     announcement
    -1.31
    enario
    -1.30
    elta
    -1.30
    bol
    -1.29
    POSITIVE LOGITS
    itch
    1.61
    ĥ½
    1.56
    TRY
    1.52
     HERE
    1.42
    necessarily
    1.41
    hing
    1.41
    ITS
    1.39
    .")
    1.38
    .");
    1.38
    urbs
    1.38
    Activations Density 0.622%

    No Known Activations