© 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. 1-RES-SM
    11. 29949
    Prev
    Next
    INDEX
    Explanations

    references to specific geographic locations or names

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    369
    +0.20
    1.1%
    410
    +0.14
    0.8%
    111
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    443
    +0.20
    0.02
    59
    +0.14
    0.02
    410
    +0.13
    0.02
    Negative Logits
     aid
    -2.20
     suitable
    -1.70
     certain
    -1.69
     help
    -1.67
     assistance
    -1.67
     different
    -1.66
    pecially
    -1.65
     best
    -1.57
     particular
    -1.56
     human
    -1.54
    POSITIVE LOGITS
    thouse
    2.03
    witz
    1.99
    leys
    1.95
    aby
    1.87
    neys
    1.84
    hurst
    1.80
    rows
    1.80
    stown
    1.79
    aughs
    1.77
    lund
    1.75
    Activations Density 0.095%

    No Known Activations