© 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. 43653
    Prev
    Next
    INDEX
    Explanations

    phrases indicating qualities or conditions of construction and fitting

    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
    ware
    -0.16
    ritch
    -0.15
    .Timeout
    -0.15
     Territory
    -0.15
     sami
    -0.15
     Pai
    -0.14
    eson
    -0.14
    TPL
    -0.14
     Territories
    -0.14
    ystals
    -0.13
    POSITIVE LOGITS
     tight
    0.58
     Tight
    0.48
     tighter
    0.48
    tight
    0.46
    -t
    0.37
     tightened
    0.37
    .tight
    0.37
     tighten
    0.36
     tightening
    0.33
     tightly
    0.31
    Activations Density 0.029%

    No Known Activations