Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 41513
    Prev
    Next
    INDEX
    Explanations

    terms describing significant events or characteristics related to legislation and technological advancements

    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
     respectively
    -0.17
    ниÑı
    -0.14
    ernity
    -0.13
    omo
    -0.13
    ä¸Ģæł·
    -0.12
    zelf
    -0.12
    ops
    -0.12
    /the
    -0.12
    asily
    -0.12
     respective
    -0.12
    POSITIVE LOGITS
    nature
    0.24
     nature
    0.24
    ness
    0.19
    人ãģ¯
    0.18
    Nature
    0.18
     Nature
    0.17
    ãģĵãģ¨ãģ¯
    0.17
     thing
    0.16
    ìŀ¥ìĿĢ
    0.15
    iveness
    0.15
    Activations Density 0.550%

    No Known Activations