© 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. 2-RES-SM
    11. 28006
    Prev
    Next
    INDEX
    Explanations

    instances of the word "for" in various 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/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
    +0.15
    0.9%
    282
    +0.14
    0.8%
    186
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.15
    0.10
    123
    +0.14
    0.10
    170
    +0.13
    0.08
    Negative Logits
    where
    -1.70
    ich
    -1.62
    cules
    -1.51
    strain
    -1.51
    page
    -1.51
    table
    -1.46
    ichi
    -1.44
    style
    -1.44
    orial
    -1.41
    guard
    -1.38
    POSITIVE LOGITS
     sale
    1.85
    ays
    1.61
     inclusion
    1.51
    cible
    1.50
     transmission
    1.47
    semin
    1.44
     clarity
    1.43
    amen
    1.41
     grabs
    1.40
    gery
    1.39
    Activations Density 0.548%

    No Known Activations