© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 24200
    Prev
    Next
    INDEX
    Explanations

    data points related to approval ratings in political 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
    ÑĢоÑĪ
    -0.18
    orna
    -0.17
    otti
    -0.15
    oss
    -0.15
     sperma
    -0.15
    iry
    -0.14
     zas
    -0.14
    hti
    -0.14
    inea
    -0.14
    lington
    -0.14
    POSITIVE LOGITS
     approval
    0.43
     Approval
    0.36
     approvals
    0.36
     ratings
    0.33
     approve
    0.33
    approval
    0.32
    Approval
    0.32
     rating
    0.30
     Ratings
    0.29
    _approval
    0.28
    Activations Density 0.051%

    No Known Activations