© 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. 44283
    Prev
    Next
    INDEX
    Explanations

    locations and positions associated with characters' actions or narratives

    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
    ibase
    -0.17
    rab
    -0.16
    atrix
    -0.16
    ngr
    -0.15
     tents
    -0.15
     Kab
    -0.15
    iba
    -0.15
     Url
    -0.14
     kab
    -0.14
     headlights
    -0.14
    POSITIVE LOGITS
     bench
    0.37
    bench
    0.29
     Bench
    0.29
     cushions
    0.28
     cushion
    0.27
     couch
    0.26
     benches
    0.25
     stool
    0.24
     Cush
    0.22
     stools
    0.20
    Activations Density 0.036%

    No Known Activations