© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 8829
    Prev
    Next
    INDEX
    Explanations

    elements of code, particularly those related to class and method definitions

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    57
    +0.12
    0.7%
    21
    +0.12
    0.7%
    123
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.12
    0.03
    245
    +0.12
    0.01
    75
    +0.11
    0.01
    Negative Logits
    aneously
    -1.73
    swers
    -1.67
    abase
    -1.67
     answer
    -1.67
    iful
    -1.66
    ronic
    -1.64
    ria
    -1.56
    ries
    -1.51
    lands
    -1.50
    hesis
    -1.49
    POSITIVE LOGITS
    µ
    4.89
    ¼
    4.87
    ½
    4.79
    ·
    4.71
    ²
    4.61
    ↵
    4.60
    <|outofrange|>
    4.60
    4.60
    č↵
    4.60
    <|outofrange|>
    4.60
    Activations Density 0.193%

    No Known Activations