© 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. 12989
    Prev
    Next
    INDEX
    Explanations

    exclamatory expressions and phrases suggesting personal engagement or feedback in a recipe context

    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
     Grund
    -0.14
     campaigns
    -0.14
    ophy
    -0.14
     Graph
    -0.13
    weets
    -0.13
     Sac
    -0.13
    _ment
    -0.13
    -backed
    -0.13
    atalog
    -0.13
     Gent
    -0.13
    POSITIVE LOGITS
     ingredients
    0.31
     ingredient
    0.29
     Ingredients
    0.28
     recipe
    0.27
    Ingredients
    0.25
     serving
    0.22
     directions
    0.21
    ingredients
    0.21
     Directions
    0.21
     Recipe
    0.21
    Activations Density 0.030%

    No Known Activations