© 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. 29047
    Prev
    Next
    INDEX
    Explanations

    prefixes and suffixes in programming contexts

    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
    278
    +0.14
    0.8%
    129
    +0.11
    0.6%
    156
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    94
    +0.14
    0.03
    278
    +0.11
    0.03
    277
    +0.11
    0.02
    Negative Logits
    rely
    -1.84
    ĥ½
    -1.68
    ILY
    -1.54
    »¿
    -1.53
    just
    -1.51
    reb
    -1.47
    ²
    -1.45
    orea
    -1.40
    ¿
    -1.39
    OND
    -1.38
    POSITIVE LOGITS
     phrase
    1.74
     notation
    1.71
    veolar
    1.69
     holders
    1.57
    smen
    1.55
     unpublished
    1.54
    idium
    1.51
     legend
    1.50
    es
    1.50
     etched
    1.49
    Activations Density 0.223%

    No Known Activations