© 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. 42972
    Prev
    Next
    INDEX
    Explanations

    phrases related to consumer awareness and informed decision-making

    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
    /setup
    -0.18
    ">//
    -0.16
    anel
    -0.15
    HAV
    -0.15
     floor
    -0.15
    urg
    -0.14
    iman
    -0.14
     Floor
    -0.14
    é§Ĩ
    -0.14
     still
    -0.14
    POSITIVE LOGITS
     dealing
    0.41
     deal
    0.35
     Deal
    0.34
     deals
    0.32
    Deal
    0.31
    deal
    0.29
     Deals
    0.28
     dealt
    0.28
     DEAL
    0.26
    -deals
    0.23
    Activations Density 0.029%

    No Known Activations