© 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 EvalsExportsinterp-engineNEWAPI 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. 25797
    Prev
    Next
    INDEX
    Explanations

    instructional phrases and steps related to using software or applications

    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
     Furn
    -0.14
    eru
    -0.14
    ihn
    -0.13
    vit
    -0.13
    edback
    -0.13
    adi
    -0.13
    à¹Ĥà¸Ĭ
    -0.12
     assignable
    -0.12
    -parser
    -0.12
    erli
    -0.12
    POSITIVE LOGITS
     how
    0.43
    how
    0.34
     usage
    0.33
    å¦Ĥä½ķ
    0.32
     basics
    0.29
     How
    0.29
     basic
    0.29
    -how
    0.28
     Usage
    0.26
     HOW
    0.25
    Activations Density 0.054%

    No Known Activations