© 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. 4-RES-SM
    11. 11188
    Prev
    Next
    INDEX
    Explanations

    instances of the word "both."

    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/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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    25
    +0.12
    0.7%
    48
    +0.11
    0.6%
    78
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    308
    +0.12
    0.03
    102
    +0.11
    0.04
    0
    +0.11
    0.03
    Negative Logits
    package
    -1.58
    riction
    -1.50
    esides
    -1.50
    skip
    -1.45
    imore
    -1.43
    ved
    -1.41
    fecture
    -1.40
     menos
    -1.38
    setlength
    -1.33
    jem
    -1.32
    POSITIVE LOGITS
     sexes
    3.73
     sides
    3.42
     halves
    2.32
     directions
    2.18
     worlds
    2.07
     chambers
    2.02
     parties
    2.02
     ends
    1.96
     kinds
    1.95
     hemisp
    1.95
    Activations Density 0.116%

    No Known Activations