© 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. 8626
    Prev
    Next
    INDEX
    Explanations

    technical specifications and measurements related to audio filters

    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
    306
    +0.25
    1.6%
    21
    +0.24
    1.5%
    322
    +0.16
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    306
    +0.25
    0.05
    21
    +0.24
    0.02
    450
    +0.16
    0.00
    Negative Logits
    ancies
    -1.58
    ilities
    -1.54
     thereof
    -1.46
    ingly
    -1.45
    \)
    -1.42
    ibilities
    -1.41
    fulness
    -1.34
    iger
    -1.32
    (\[
    -1.30
    \.
    -1.29
    POSITIVE LOGITS
    ĻĤ
    2.38
    ¢
    2.02
    IJ
    2.01
    ¿
    1.99
    Īĺ
    1.96
    Ī
    1.89
    Ł
    1.89
    Ĥ¬
    1.87
    Ĺ
    1.86
    °
    1.86
    Activations Density 3.553%

    No Known Activations