© 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. 1-RES-SM
    11. 29483
    Prev
    Next
    INDEX
    Explanations

    the occurrence of the word "da" 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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    258
    +0.14
    0.8%
    369
    +0.13
    0.8%
    241
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.14
    0.03
    271
    +0.13
    0.00
    156
    +0.11
    0.02
    Negative Logits
    ĩ
    -1.87
    ĵ
    -1.80
    Ĺ
    -1.72
     apologize
    -1.69
    Īĺ
    -1.68
    Ĥ
    -1.67
    ij
    -1.67
    ¿½
    -1.62
    ħ
    -1.62
    Ĩ
    -1.62
    POSITIVE LOGITS
    emon
    1.90
    hedral
    1.77
    uncher
    1.67
    ÅĽci
    1.67
    eton
    1.61
    fulness
    1.56
    istics
    1.54
    place
    1.54
    amento
    1.54
     Vinc
    1.53
    Activations Density 0.191%

    No Known Activations