© 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. 1-RES-SM
    11. 24388
    Prev
    Next
    INDEX
    Explanations

    instances of emphasis or strong feelings expressed in text

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    478
    +0.13
    0.7%
    355
    +0.12
    0.7%
    219
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    355
    +0.13
    0.09
    441
    +0.12
    0.04
    111
    +0.12
    0.06
    Negative Logits
    )|$(
    -1.87
    ĩ
    -1.82
    ©
    -1.61
    ¢
    -1.60
    '?
    -1.55
     ':
    -1.46
     times
    -1.45
    )?
    -1.44
    lings
    -1.39
    ĸ´
    -1.37
    POSITIVE LOGITS
    anas
    1.60
    ves
    1.46
    tain
    1.46
    usement
    1.43
    pless
    1.41
    uts
    1.41
     appeal
    1.40
    tu
    1.39
    ses
    1.39
    amily
    1.37
    Activations Density 0.599%

    No Known Activations