© 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. 8740
    Prev
    Next
    INDEX
    Explanations

    references to the concept of "home."

    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
    246
    +0.17
    0.9%
    78
    +0.14
    0.8%
    26
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    246
    +0.17
    0.05
    489
    +0.14
    0.03
    20
    +0.12
    0.04
    Negative Logits
    ing
    -2.48
    ¯
    -1.97
    hetti
    -1.78
    ingen
    -1.70
    Ŀ
    -1.68
    pire
    -1.62
    ftware
    -1.61
    tle
    -1.49
    nell
    -1.43
    ĥ
    -1.43
    POSITIVE LOGITS
    owner
    3.28
    grown
    2.98
    brew
    2.89
    comings
    2.82
    coming
    2.78
    owners
    2.39
    own
    2.32
    opathic
    2.21
    buy
    2.17
    ostatic
    2.16
    Activations Density 0.262%

    No Known Activations