© 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. 14453
    Prev
    Next
    INDEX
    Explanations

    references to copyright and related legal concepts

    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
    271
    +0.16
    1.0%
    78
    +0.13
    0.8%
    160
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    460
    +0.16
    0.01
    78
    +0.13
    -0.00
    255
    +0.11
    -0.00
    Negative Logits
    hin
    -1.80
    ities
    -1.57
    abouts
    -1.44
    ctic
    -1.39
    gio
    -1.38
    merce
    -1.30
    gi
    -1.28
    defined
    -1.25
    inities
    -1.24
    nament
    -1.24
    POSITIVE LOGITS
    áĢº
    1.93
     struggle
    1.60
    à§ĩ
    1.38
    á½IJ
    1.38
     embarrassing
    1.37
    onsense
    1.33
    owl
    1.32
    esty
    1.31
    áĢ¯
    1.31
    áĢ
    1.28
    Activations Density 0.095%

    No Known Activations