© 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. 10417
    Prev
    Next
    INDEX
    Explanations

    programming commands and their parameters

    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
    +0.17
    1.0%
    78
    +0.13
    0.7%
    32
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    148
    +0.17
    0.02
    444
    +0.13
    0.04
    157
    +0.11
    0.03
    Negative Logits
    tum
    -1.53
    alle
    -1.51
    iates
    -1.50
    ieur
    -1.45
    edin
    -1.45
    leans
    -1.45
    arily
    -1.42
     chemistry
    -1.42
    iella
    -1.40
    iative
    -1.39
    POSITIVE LOGITS
    Į
    2.53
    Ļª
    2.49
    į
    2.20
    ·
    2.15
    º
    2.10
    ¤
    1.97
     '*
    1.92
     "$
    1.91
     "$(
    1.91
    ¹
    1.89
    Activations Density 0.318%

    No Known Activations