© 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. 24563
    Prev
    Next
    INDEX
    Explanations

    instructions related to software installation and management

    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
    aus
    -0.15
    ↵
    -0.15
    owing
    -0.14
    oth
    -0.13
    Portrait
    -0.13
    inha
    -0.13
    quiring
    -0.12
    aggi
    -0.12
    erken
    -0.12
    Ùģس
    -0.12
    POSITIVE LOGITS
     once
    0.36
    once
    0.32
     Once
    0.31
    Once
    0.29
     now
    0.29
     Now
    0.27
    now
    0.26
    Now
    0.24
     after
    0.24
     Finally
    0.23
    Activations Density 0.105%

    No Known Activations