© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 14539
    Prev
    Next
    INDEX
    Explanations

    instances of the word "lo" in various forms, indicating a focus on exclamations or expressions of surprise or admiration

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    406
    +0.16
    0.9%
    78
    +0.14
    0.8%
    109
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    406
    +0.16
    0.03
    109
    +0.14
    0.03
    31
    +0.13
    0.02
    Negative Logits
    ments
    -1.77
    ment
    -1.73
    line
    -1.65
    unnumbered
    -1.62
    nement
    -1.62
    manship
    -1.57
    lines
    -1.52
    lace
    -1.52
    mith
    -1.50
    liest
    -1.50
    POSITIVE LOGITS
    ĭ
    3.74
    Ľ
    3.69
    ·¸
    3.42
    Ħ
    3.42
    Ŀ
    3.42
    ī
    3.37
    Ĩ
    3.36
    IJ
    3.34
    ¶
    3.33
    ĸ´
    3.33
    Activations Density 0.154%

    No Known Activations