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

    Neuronpedia

    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. 40753
    Prev
    Next
    INDEX
    Explanations

    phrases that describe the presence of features or attributes in objects or components

    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
    aken
    -0.16
    lick
    -0.14
    yc
    -0.14
    ีà¸ļ
    -0.13
    æľīä»Ģä¹Ī
    -0.13
    ean
    -0.13
    ordinate
    -0.13
    ime
    -0.13
    ningen
    -0.13
    itz
    -0.13
    POSITIVE LOGITS
    stood
    0.35
     attached
    0.27
    outh
    0.26
     accompanying
    0.25
    standing
    0.24
     regard
    0.23
    ered
    0.22
     associated
    0.22
     embedded
    0.21
     added
    0.21
    Activations Density 0.184%

    No Known Activations