© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 28910
    Prev
    Next
    INDEX
    Explanations

    function return statements in programming code

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    478
    +0.19
    1.0%
    240
    +0.11
    0.6%
    199
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    299
    +0.19
    0.03
    206
    +0.11
    0.02
    113
    +0.10
    0.04
    Negative Logits
    ther
    -1.81
    orro
    -1.81
    sburg
    -1.76
    amon
    -1.76
    ero
    -1.74
    rug
    -1.71
    )\]
    -1.66
    ctomy
    -1.63
    oin
    -1.62
    utz
    -1.62
    POSITIVE LOGITS
    µ
    1.66
     favour
    1.65
     quo
    1.62
     yes
    1.62
     parl
    1.58
     things
    1.57
     false
    1.56
     matern
    1.55
     clear
    1.55
     conventional
    1.54
    Activations Density 0.162%

    No Known Activations