© 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. 8557
    Prev
    Next
    INDEX
    Explanations

    patterns of structured data or code snippets

    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
    126
    +0.18
    1.0%
    78
    +0.16
    0.9%
    255
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    141
    +0.18
    0.04
    123
    +0.16
    0.04
    241
    +0.14
    0.04
    Negative Logits
    amer
    -1.60
    wise
    -1.51
    arious
    -1.43
    utors
    -1.39
    oths
    -1.35
     elegans
    -1.32
     birds
    -1.31
     blacks
    -1.30
    ony
    -1.30
    ons
    -1.28
    POSITIVE LOGITS
    ®
    2.10
    Ĥ¬
    1.99
    ·
    1.99
    Ĩ
    1.93
    ³
    1.88
    ¦
    1.86
    ĩ
    1.80
    à±į
    1.76
    Ļ
    1.75
    «
    1.75
    Activations Density 0.271%

    No Known Activations