© 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. 4-RES-SM
    11. 10370
    Prev
    Next
    INDEX
    Explanations

    expressions of gratitude and appreciation 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/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
    348
    +0.17
    1.0%
    306
    +0.16
    0.9%
    509
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    366
    +0.17
    0.07
    390
    +0.16
    0.13
    478
    +0.12
    0.04
    Negative Logits
    icolor
    -1.53
    pole
    -1.48
    ?,
    -1.47
    quart
    -1.36
    spray
    -1.35
     (),
    -1.34
     affairs
    -1.30
     thousands
    -1.29
    uler
    -1.29
     Photograph
    -1.28
    POSITIVE LOGITS
     note
    2.42
     hope
    2.29
     regards
    2.03
     appreciated
    2.01
     thank
    1.98
     Conclusion
    1.97
     future
    1.96
    idenote
    1.95
     thanks
    1.95
     hopefully
    1.95
    Activations Density 1.987%

    No Known Activations