© 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. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 42585
    Prev
    Next
    INDEX
    Explanations

    terms related to finalization or completion

    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
    htags
    -0.20
    ated
    -0.17
    ified
    -0.15
    oretical
    -0.14
    rame
    -0.14
    noinspection
    -0.14
    smouth
    -0.14
    ToPoint
    -0.14
    rait
    -0.14
    еÑĤиÑĩ
    -0.14
    POSITIVE LOGITS
    izing
    0.38
    ised
    0.35
    ising
    0.33
    izes
    0.33
    ize
    0.28
    y
    0.28
    mente
    0.27
    ises
    0.26
    istically
    0.26
    izable
    0.26
    Activations Density 0.001%

    No Known Activations