© 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. 26691
    Prev
    Next
    INDEX
    Explanations

    instances of fraudulent email schemes

    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
    426
    +0.15
    0.8%
    26
    +0.14
    0.8%
    360
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    12
    +0.15
    0.02
    26
    +0.14
    0.01
    360
    +0.12
    0.02
    Negative Logits
     definite
    -1.58
     weekends
    -1.56
     believing
    -1.53
     feeling
    -1.47
     Independence
    -1.47
    respectively
    -1.47
     SO
    -1.46
     Afterwards
    -1.41
     holidays
    -1.41
    cock
    -1.40
    POSITIVE LOGITS
    »
    2.95
    Ģ
    2.71
    Ĥ¬
    2.68
    ĨĴ
    2.68
    Ĵ
    2.61
    Ļª
    2.55
    ¾
    2.53
    Ń
    2.53
    2.50
    ↵
    2.50
    Activations Density 0.228%

    No Known Activations