© 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. 13752
    Prev
    Next
    INDEX
    Explanations

    expressions of hope and optimism

    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
    15
    +0.15
    0.9%
    428
    +0.11
    0.6%
    232
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.15
    0.07
    106
    +0.11
    0.04
    56
    +0.11
    0.03
    Negative Logits
    £
    -3.81
    «
    -3.66
    °
    -3.53
    ¿
    -3.51
    į
    -3.49
    Ŀ
    -3.46
    ĥ½
    -3.44
    ¢
    -3.39
    ĥ
    -3.39
    ·
    -3.37
    POSITIVE LOGITS
    fully
    2.90
    ful
    2.31
     someone
    1.94
    lessly
    1.86
     somebody
    1.82
    full
    1.82
     someday
    1.72
     they
    1.71
     that
    1.71
     you
    1.68
    Activations Density 0.490%

    No Known Activations