© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 33366
    Prev
    Next
    INDEX
    Explanations

    terminology related to factories and industrial activities

    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
     agencies
    -0.15
     agency
    -0.15
    ï¸
    -0.15
     кÑĥлÑĮÑĤÑĥ
    -0.14
    ĵĺ
    -0.14
    à¸Ńà¸ĩà¸Ħ
    -0.14
    952
    -0.14
    ServiceProvider
    -0.13
    adro
    -0.13
    agency
    -0.13
    POSITIVE LOGITS
     producing
    0.27
     output
    0.22
     operation
    0.22
     located
    0.21
     operations
    0.21
     production
    0.21
     processing
    0.20
     produces
    0.20
     Produ
    0.20
     produce
    0.20
    Activations Density 0.038%

    No Known Activations