© 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. 43594
    Prev
    Next
    INDEX
    Explanations

    terms related to bureaucratic processes and infrastructure project management

    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
    gars
    -0.16
    beck
    -0.15
     grav
    -0.15
    ayah
    -0.13
     crest
    -0.13
    eguard
    -0.13
     distur
    -0.13
    æµħ
    -0.13
    isd
    -0.13
     panicked
    -0.13
    POSITIVE LOGITS
     bureaucracy
    0.37
     bureaucratic
    0.36
     paperwork
    0.31
     bureauc
    0.28
     red
    0.28
     administrative
    0.25
     Paper
    0.25
     bü
    0.25
     hoops
    0.24
     paper
    0.23
    Activations Density 0.060%

    No Known Activations