© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 25337
    Prev
    Next
    INDEX
    Explanations

    occurrences of the definite article "la" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    hausen
    -0.17
     escorte
    -0.16
     nackte
    -0.16
    contri
    -0.15
    ga
    -0.15
    thon
    -0.14
     labore
    -0.14
    chan
    -0.14
    Å«
    -0.14
    b
    -0.14
    POSITIVE LOGITS
    unched
    0.24
    uded
    0.19
    ufen
    0.18
    ids
    0.17
     same
    0.17
     tend
    0.17
    zos
    0.17
    urence
    0.16
    isser
    0.16
    uds
    0.16
    Activations Density 0.008%

    No Known Activations