© 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 EvalsExportsAPI 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. 26604
    Prev
    Next
    INDEX
    Explanations

    instances of the word "primary."

    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
    340
    +0.12
    0.7%
    3
    +0.12
    0.6%
    343
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    3
    +0.12
    0.02
    235
    +0.12
    0.02
    340
    +0.11
    0.02
    Negative Logits
    ĥ½
    -1.99
    Ĵ
    -1.95
    ¡
    -1.88
    ¥
    -1.87
    ½
    -1.81
    Ī
    -1.69
    ť
    -1.64
    §
    -1.62
     latter
    -1.61
    ľ
    -1.60
    POSITIVE LOGITS
    hole
    1.81
    esian
    1.59
    ibilities
    1.59
     generations
    1.53
    holes
    1.52
    prints
    1.47
    quarters
    1.46
     duties
    1.43
    hog
    1.43
    ised
    1.42
    Activations Density 0.038%

    No Known Activations