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. 23521
    Prev
    Next
    INDEX
    Explanations

    references to political threats and security concerns

    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
    pper
    -0.16
    xygen
    -0.15
    antz
    -0.15
    ÐĴС
    -0.14
    αÏģά
    -0.14
    Thunk
    -0.14
    .Prot
    -0.13
    ervas
    -0.13
     Ù쨶
    -0.13
    assi
    -0.13
    POSITIVE LOGITS
     pose
    0.87
     poses
    0.80
     posing
    0.70
     Pose
    0.70
     posed
    0.70
    pose
    0.66
    Pose
    0.64
    poses
    0.62
     presents
    0.62
     presenting
    0.53
    Activations Density 0.054%

    No Known Activations