© 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. 12979
    Prev
    Next
    INDEX
    Explanations

    references to the term "ter" 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
    98
    +0.12
    0.7%
    431
    +0.12
    0.7%
    352
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.12
    0.02
    431
    +0.12
    0.02
    351
    +0.11
    0.02
    Negative Logits
     wont
    -1.50
     kidding
    -1.47
     exception
    -1.42
    hered
    -1.41
     scratch
    -1.41
     sender
    -1.39
    quet
    -1.36
     copyrighted
    -1.35
    cgi
    -1.32
     doubt
    -1.29
    POSITIVE LOGITS
    rors
    2.02
    neys
    1.88
    ogeneity
    1.82
    iors
    1.79
    tainment
    1.75
    prises
    1.75
    tes
    1.72
    velt
    1.67
    illas
    1.65
    ilda
    1.59
    Activations Density 0.199%

    No Known Activations