© 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. 14437
    Prev
    Next
    INDEX
    Explanations

    instances of the word "to" in various contexts

    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
     choice
    -0.33
     Choice
    -0.28
     choices
    -0.25
    choice
    -0.24
     choix
    -0.24
    -choice
    -0.23
    _choice
    -0.21
     choose
    -0.20
    Choice
    -0.20
    _choices
    -0.19
    POSITIVE LOGITS
    ASM
    0.16
    stride
    0.15
    idebar
    0.15
    _COMPAT
    0.15
    .lu
    0.15
    entially
    0.15
    çĽ¸ä¿¡
    0.14
    ectar
    0.14
    ady
    0.14
    acky
    0.14
    Activations Density 0.016%

    No Known Activations