© 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. 15122
    Prev
    Next
    INDEX
    Explanations

    numerical values and their associated contextual significance

    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
    69
    +0.11
    0.6%
    206
    +0.11
    0.6%
    306
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.11
    0.06
    221
    +0.11
    0.04
    150
    +0.11
    0.05
    Negative Logits
    ner
    -1.66
    ily
    -1.63
    ettes
    -1.55
    ette
    -1.54
    ify
    -1.50
     dated
    -1.50
    ified
    -1.48
    horn
    -1.47
    izer
    -1.46
    iform
    -1.46
    POSITIVE LOGITS
     least
    2.81
    mosp
    2.22
     intervals
    2.18
    rophy
    2.02
    least
    1.83
    las
    1.82
     various
    1.76
     distances
    1.75
     low
    1.74
     once
    1.74
    Activations Density 0.133%

    No Known Activations