© 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. 9710
    Prev
    Next
    INDEX
    Explanations

    references and citations

    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
    480
    +0.12
    0.6%
    225
    +0.11
    0.6%
    348
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    401
    +0.12
    0.01
    394
    +0.11
    0.02
    31
    +0.11
    0.02
    Negative Logits
    ¯
    -2.14
    ĥ½
    -1.86
    akin
    -1.69
    xico
    -1.61
    mable
    -1.61
    º
    -1.58
    inez
    -1.56
    ľĵ
    -1.50
    iej
    -1.49
    arks
    -1.47
    POSITIVE LOGITS
     thereto
    2.27
     point
    1.97
     material
    1.87
    frame
    1.83
     points
    1.73
     to
    1.72
     frame
    1.71
     frames
    1.68
     numer
    1.65
     lists
    1.62
    Activations Density 0.306%

    No Known Activations