© 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. 14001
    Prev
    Next
    INDEX
    Explanations

    the conjunction "and" to identify connections or additions in sentences

    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
     annonces
    -0.14
    év
    -0.13
     hã
    -0.13
    ạt
    -0.12
    ãn
    -0.12
    âĢį
    -0.12
    iets
    -0.12
    ocations
    -0.12
    aginator
    -0.12
    zano
    -0.12
    POSITIVE LOGITS
    /or
    0.49
    erson
    0.26
    rog
    0.26
    alus
    0.25
    rogen
    0.24
    ÑĢониÑĩеÑģ
    0.23
    /OR
    0.23
    rew
    0.22
    наÑĩе
    0.21
    hra
    0.20
    Activations Density 0.294%

    No Known Activations