© 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. 4-RES-SM
    11. 13297
    Prev
    Next
    INDEX
    Explanations

    names of geographical regions and locations

    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
    116
    +0.14
    0.8%
    326
    +0.14
    0.8%
    52
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    116
    +0.14
    0.04
    52
    +0.14
    0.04
    326
    +0.13
    0.04
    Negative Logits
    ĨĴ
    -5.32
    Ļª
    -5.11
    £
    -5.09
    Ļ
    -5.02
    į
    -4.90
    ŀ
    -4.86
    ¢
    -4.78
    ¦
    -4.74
    Īĺ
    -4.72
    «
    -4.71
    POSITIVE LOGITS
    shire
    2.62
     Province
    2.29
    ensis
    2.27
     forests
    2.27
     soils
    2.10
     Territory
    2.03
     coast
    2.01
     cattle
    2.00
     Islands
    1.99
     waters
    1.97
    Activations Density 0.262%

    No Known Activations