© 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. 12973
    Prev
    Next
    INDEX
    Explanations

    symbols or punctuation marks, particularly colons

    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
    156
    +0.48
    3.1%
    478
    +0.21
    1.4%
    271
    +0.17
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    150
    +0.48
    0.02
    161
    +0.21
    0.01
    476
    +0.17
    0.01
    Negative Logits
    liest
    -1.53
    ously
    -1.41
    thood
    -1.33
    icago
    -1.33
     orders
    -1.32
    ipel
    -1.32
    poons
    -1.28
     versions
    -1.27
    lier
    -1.26
    otle
    -1.25
    POSITIVE LOGITS
    ·
    1.60
    »
    1.55
    mozilla
    1.55
    ¯
    1.51
    ©
    1.49
    ĻĤ
    1.39
    ¼
    1.38
    defer
    1.38
    ĸ´
    1.34
    ¤
    1.34
    Activations Density 0.025%

    No Known Activations