© 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. 22360
    Prev
    Next
    INDEX
    Explanations

    questions and discussions about marriage and relationships

    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
     Gee
    -0.16
    ellow
    -0.15
     thiên
    -0.15
     Unused
    -0.15
    íĹĪ
    -0.14
    illa
    -0.14
     rtrim
    -0.14
    arez
    -0.14
    ivar
    -0.13
     earnest
    -0.13
    POSITIVE LOGITS
     forget
    0.44
    forget
    0.39
     Forget
    0.38
    Forget
    0.35
     tough
    0.33
     Tough
    0.32
     screw
    0.26
     forgetting
    0.25
     Sorry
    0.24
     toughness
    0.24
    Activations Density 0.048%

    No Known Activations