© 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. 2-RES-SM
    11. 32283
    Prev
    Next
    INDEX
    Explanations

    references to newspapers and publications

    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/2-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.2.hook_resid_post
    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
    362
    +0.14
    0.8%
    156
    +0.12
    0.6%
    38
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    38
    +0.14
    0.01
    64
    +0.12
    0.02
    45
    +0.11
    0.01
    Negative Logits
    ograft
    -1.67
    owed
    -1.64
    ouin
    -1.63
    uetooth
    -1.54
     autophagy
    -1.52
    ĥ½
    -1.49
    chemia
    -1.46
     autism
    -1.46
    asio
    -1.45
     dissociation
    -1.43
    POSITIVE LOGITS
    matic
    1.76
    holders
    1.72
     Bomb
    1.50
    inals
    1.48
     Bengal
    1.44
     coverage
    1.43
     headlines
    1.41
     cart
    1.40
    manship
    1.40
    ized
    1.39
    Activations Density 0.276%

    No Known Activations