© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 14601
    Prev
    Next
    INDEX
    Explanations

    actions and processes related to research and analysis in academic writing

    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
     exactly
    -0.15
    ĶåĽŀ
    -0.14
    what
    -0.14
     nÃło
    -0.14
    minated
    -0.14
     what
    -0.14
    iken
    -0.14
    wich
    -0.14
     something
    -0.13
    something
    -0.13
    POSITIVE LOGITS
     these
    0.16
    ýt
    0.14
    .UnitTesting
    0.13
    strncmp
    0.13
    è¿ĻäºĽ
    0.13
    atte
    0.13
    eÅŁ
    0.12
    ÏĨÏħ
    0.12
    annt
    0.12
    isFunction
    0.12
    Activations Density 0.037%

    No Known Activations