© 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 EvalsExportsinterp-engineNEWAPI 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. 26552
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "sat."

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    376
    +0.14
    0.8%
    148
    +0.14
    0.8%
    218
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    112
    +0.14
    0.02
    218
    +0.14
    0.02
    451
    +0.11
    0.02
    Negative Logits
    ments
    -1.73
     bulk
    -1.65
     lig
    -1.40
     attention
    -1.37
     commercial
    -1.35
     skin
    -1.35
    wards
    -1.34
     capsule
    -1.34
     distribution
    -1.30
    icin
    -1.29
    POSITIVE LOGITS
    irical
    2.27
     atop
    1.76
    iated
    1.73
    among
    1.66
    riz
    1.64
     aloud
    1.63
    iful
    1.60
    uated
    1.58
     squarely
    1.53
    yrinth
    1.53
    Activations Density 0.009%

    No Known Activations