© 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. 54990
    Prev
    Next
    INDEX
    Explanations

    mention of specific U.S. states

    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
     Byl
    -0.15
    ij¸
    -0.14
    ÑĥÑĤи
    -0.14
    ffc
    -0.13
    еÑĢж
    -0.13
    azzi
    -0.13
    owy
    -0.13
    YW
    -0.12
    -vs
    -0.12
    est
    -0.12
    POSITIVE LOGITS
     USA
    0.23
    USA
    0.19
    -area
    0.18
     area
    0.17
    achuset
    0.15
     Territory
    0.15
    Area
    0.14
    «ĺ
    0.14
     área
    0.14
     hometown
    0.14
    Activations Density 0.053%

    No Known Activations