Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 31292
    Prev
    Next
    INDEX
    Explanations

    instances of criticism regarding laws or procedures related to privacy and civil liberties

    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
    éĽĦ
    -0.14
    zap
    -0.14
     Poll
    -0.14
    _poll
    -0.14
    .Parameters
    -0.14
     Calcium
    -0.14
    ouncer
    -0.14
    uture
    -0.13
    Poll
    -0.13
    åŀ
    -0.13
    POSITIVE LOGITS
     civil
    0.39
     groups
    0.33
     critics
    0.32
     Civil
    0.28
     Groups
    0.28
     Critics
    0.26
     opponents
    0.26
    Crit
    0.25
    Civil
    0.25
     rights
    0.25
    Activations Density 0.064%

    No Known Activations