© 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. 29035
    Prev
    Next
    INDEX
    Explanations

    special character patterns and formatting in code 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/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
    12
    +0.15
    0.8%
    466
    +0.11
    0.6%
    53
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    12
    +0.15
    0.03
    136
    +0.11
    -0.00
    151
    +0.11
    0.03
    Negative Logits
    ľ
    -2.28
    ĩ
    -2.20
    ´
    -2.14
    ¿½
    -2.08
    ¯
    -1.99
    ŀ
    -1.99
    Ĭ
    -1.99
    boarding
    -1.92
    ¬
    -1.91
    ĸ´
    -1.83
    POSITIVE LOGITS
    room
    1.55
    IRE
    1.47
    ETHOD
    1.44
    ("%
    1.44
     Research
    1.38
    ought
    1.38
    agon
    1.36
    rooms
    1.34
    lee
    1.32
    vre
    1.32
    Activations Density 0.526%

    No Known Activations