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. 23542
    Prev
    Next
    INDEX
    Explanations

    safety precautions and recommendations for furniture stability and fire prevention

    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
     myself
    -0.12
    istik
    -0.12
    OPTIONS
    -0.12
    iko
    -0.12
    urar
    -0.12
     HELP
    -0.11
    Interest
    -0.11
    angu
    -0.11
     ourselves
    -0.11
    Interested
    -0.11
    POSITIVE LOGITS
     always
    0.38
     Always
    0.34
     avoid
    0.31
     never
    0.30
    Always
    0.30
    always
    0.30
     ALWAYS
    0.28
     Avoid
    0.27
     NEVER
    0.27
     remember
    0.27
    Activations Density 0.178%

    No Known Activations