© 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. 33565
    Prev
    Next
    INDEX
    Explanations

    references to energy generation and infrastructure

    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
    xes
    -0.14
     cÃŃt
    -0.14
    rung
    -0.13
    usp
    -0.13
    ÑĢик
    -0.13
     ðŁĺī↵↵
    -0.13
    stal
    -0.13
     दर
    -0.13
    etzt
    -0.12
    CUS
    -0.12
    POSITIVE LOGITS
     cost
    0.28
    cost
    0.24
     costs
    0.23
    -cost
    0.22
     technology
    0.21
     Cost
    0.21
     practical
    0.21
     COST
    0.21
    Cost
    0.20
     technical
    0.20
    Activations Density 0.106%

    No Known Activations