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. 31288
    Prev
    Next
    INDEX
    Explanations

    expressions of inadequate recognition or appreciation, particularly towards struggles or contributions

    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
    inois
    -0.16
    354
    -0.15
    intree
    -0.14
    _FINE
    -0.14
    ooth
    -0.14
    ĥ½
    -0.14
    GV
    -0.14
    ocking
    -0.13
    енÑģ
    -0.13
     ë°Ķë¡ľ
    -0.13
    POSITIVE LOGITS
     nearly
    0.77
     Nearly
    0.66
    Nearly
    0.57
     anywhere
    0.54
     near
    0.40
     remotely
    0.39
     half
    0.36
     Near
    0.35
     nowhere
    0.33
     HALF
    0.33
    Activations Density 0.052%

    No Known Activations