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. 23451
    Prev
    Next
    INDEX
    Explanations

    elements related to the supernatural or haunted themes

    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
    ayne
    -0.14
    олог
    -0.14
    #Region
    -0.14
    åħ¬åijĬ
    -0.14
    .EventSystems
    -0.13
    лиз
    -0.13
    άλι
    -0.13
    }\.[
    -0.13
    andle
    -0.13
    FK
    -0.12
    POSITIVE LOGITS
     army
    0.72
     armies
    0.57
     Army
    0.51
    Army
    0.47
     hor
    0.47
     fleet
    0.44
     legion
    0.43
     leg
    0.42
     fleets
    0.38
    arm
    0.38
    Activations Density 0.045%

    No Known Activations