© 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. 14406
    Prev
    Next
    INDEX
    Explanations

    technical terms and parameters related to data processing or configuration settings

    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
    181
    +0.13
    0.7%
    50
    +0.12
    0.7%
    97
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    338
    +0.13
    0.03
    97
    +0.12
    0.05
    181
    +0.11
    0.04
    Negative Logits
    LECT
    -1.96
     sorry
    -1.50
    ![
    -1.47
    EDIT
    -1.43
    ======
    -1.41
    ****
    -1.38
    wikipedia
    -1.36
    SO
    -1.36
    idenote
    -1.35
     ABOUT
    -1.32
    POSITIVE LOGITS
    ariate
    1.88
    áĢº
    1.63
    yset
    1.62
    cht
    1.52
     Vine
    1.44
    ager
    1.43
    alia
    1.42
    ¾
    1.42
    thood
    1.40
     Wine
    1.39
    Activations Density 0.374%

    No Known Activations