© 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. 2-RES-SM
    11. 29781
    Prev
    Next
    INDEX
    Explanations

    references to parts or sections in a document or text

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    371
    +0.14
    0.8%
    502
    +0.14
    0.8%
    62
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    371
    +0.14
    0.04
    343
    +0.14
    0.04
    62
    +0.13
    0.03
    Negative Logits
    ¢
    -2.80
    Īĺ
    -2.56
    ľĵ
    -2.46
    ĻĤ
    -2.45
    ĨĴ
    -2.41
    ¡
    -2.38
    ¦
    -2.32
    ¥
    -2.23
    ĵ
    -2.15
    Ń
    -2.11
    POSITIVE LOGITS
    isans
    2.58
    ridge
    2.42
    ridges
    2.25
     of
    2.01
     thereof
    1.83
    icularly
    1.79
    ners
    1.78
    icular
    1.75
    meal
    1.75
    nered
    1.75
    Activations Density 0.070%

    No Known Activations