© 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. 23757
    Prev
    Next
    INDEX
    Explanations

    instances of the word "even" indicating emphasis or contrast

    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
    ersistent
    -0.15
    à¹ĥà¸Ķ
    -0.15
    /Gate
    -0.15
    ByExample
    -0.14
    umd
    -0.14
    osloven
    -0.14
    обÑĢеÑĤ
    -0.13
    senal
    -0.13
    å¨ĺ
    -0.13
    optic
    -0.13
    POSITIVE LOGITS
     though
    0.50
    though
    0.39
    Though
    0.38
     Though
    0.37
     aunque
    0.27
    ings
    0.25
    iment
    0.22
     tho
    0.22
     with
    0.22
    ness
    0.21
    Activations Density 0.045%

    No Known Activations