© 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. 12629
    Prev
    Next
    INDEX
    Explanations

    the word "however" signifying contrasting ideas or conditions

    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
    58
    +0.10
    0.6%
    274
    +0.10
    0.5%
    127
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    58
    +0.10
    0.04
    200
    +0.10
    0.04
    221
    +0.10
    0.03
    Negative Logits
     wonders
    -1.70
    ennis
    -1.61
    ale
    -1.59
     minds
    -1.49
    illin
    -1.48
     together
    -1.47
    alent
    -1.46
     deserves
    -1.46
    /">
    -1.34
     example
    -1.33
    POSITIVE LOGITS
    ³
    1.47
     Ground
    1.47
    fered
    1.38
    imal
    1.33
    ł
    1.33
    sein
    1.29
     overriding
    1.28
     Determ
    1.28
     Post
    1.28
    gni
    1.27
    Activations Density 0.095%

    No Known Activations