© 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. 25301
    Prev
    Next
    INDEX
    Explanations

    numerical values and indicators of time or history

    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
    ora
    -0.17
    ORA
    -0.16
    ÏĥÏĦα
    -0.15
     tonight
    -0.15
     tomorrow
    -0.14
     Tonight
    -0.14
    ãĥ¼ãĥĹ
    -0.14
    ãĥ¼ãĥģ
    -0.14
    Ãłn
    -0.14
    Tonight
    -0.13
    POSITIVE LOGITS
    ish
    0.43
    ISH
    0.31
     or
    0.29
    -ish
    0.24
     oder
    0.24
    å·¦åı³
    0.24
    辺
    0.23
     maybe
    0.23
    æĪĸèĢħ
    0.22
    æĪĸ
    0.22
    Activations Density 0.056%

    No Known Activations