© 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 EvalsExportsinterp-engineNEWAPI 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. 24328
    Prev
    Next
    INDEX
    Explanations

    commas in the 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/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
    499
    +0.15
    0.9%
    391
    +0.12
    0.7%
    203
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    391
    +0.15
    0.03
    203
    +0.12
    0.03
    499
    +0.11
    0.03
    Negative Logits
    ĻĤ
    -2.57
    ľĵ
    -2.16
    ¿½
    -2.05
    »
    -1.93
    ¦
    -1.93
    ¢
    -1.83
    Ĥ
    -1.77
    IJ
    -1.76
    Ĩ
    -1.75
    Ł
    -1.75
    POSITIVE LOGITS
    imum
    1.56
    sein
    1.45
     modulus
    1.34
     always
    1.28
     derivative
    1.25
    aq
    1.24
     dass
    1.24
     born
    1.24
    undo
    1.23
     sever
    1.23
    Activations Density 0.303%

    No Known Activations