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

    Neuronpedia

    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. 27158
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "White."

    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
    369
    +0.15
    0.9%
    362
    +0.13
    0.7%
    443
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    477
    +0.15
    0.04
    443
    +0.13
    0.04
    362
    +0.12
    0.04
    Negative Logits
    ¯
    -3.15
    ĵ
    -2.81
    ľĵ
    -2.63
    Ĥ
    -2.60
    Ĺ
    -2.56
    IJ
    -2.53
    ĥ½
    -2.47
    ľ
    -2.46
    Ń
    -2.40
    ©
    -2.40
    POSITIVE LOGITS
    ford
    2.38
    holm
    2.11
    hurst
    2.08
    heads
    1.98
    stown
    1.87
    stock
    1.86
    xton
    1.85
    igan
    1.85
    ños
    1.81
    gate
    1.81
    Activations Density 0.193%

    No Known Activations