© 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. Residuals
    9. 4-RES-SM
    11. 8571
    Prev
    Next
    INDEX
    Explanations

    terminology related to health and disease prevention

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/4-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.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    333
    +0.14
    0.8%
    125
    +0.12
    0.7%
    63
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    63
    +0.14
    0.01
    333
    +0.12
    0.04
    186
    +0.12
    0.03
    Negative Logits
    minster
    -2.16
     Ky
    -2.01
     Navigation
    -1.81
     Mason
    -1.79
     Fry
    -1.79
     navigation
    -1.77
     paddle
    -1.74
     Geometry
    -1.70
     Potter
    -1.66
     Bark
    -1.66
    POSITIVE LOGITS
     HIV
    4.83
    HIV
    4.53
     AIDS
    4.03
    AIDS
    3.54
     HCV
    3.43
     infected
    3.24
     immunodeficiency
    3.05
     viral
    3.02
     ART
    2.95
     infect
    2.93
    Activations Density 0.543%

    No Known Activations