© 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. 13250
    Prev
    Next
    INDEX
    Explanations

    adverbs that qualify or elaborate on actions or statements

    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
    307
    +0.33
    2.1%
    306
    +0.15
    1.0%
    274
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.33
    0.08
    458
    +0.15
    0.04
    316
    +0.13
    0.07
    Negative Logits
    ħ
    -2.29
    ŀ
    -2.11
    Ħ
    -2.08
    ĥ
    -2.04
    Ģ
    -2.03
    Ļª
    -1.99
    ľĵ
    -1.91
    º
    -1.89
    ĭ
    -1.88
    İ
    -1.83
    POSITIVE LOGITS
     there
    1.99
     however
    1.86
     these
    1.68
     speaking
    1.65
     it
    1.61
    there
    1.56
     though
    1.54
     they
    1.51
     none
    1.50
     choosing
    1.49
    Activations Density 0.504%

    No Known Activations