© 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. 30419
    Prev
    Next
    INDEX
    Explanations

    the word "make" in various contexts, indicating creation or production

    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
    mos
    -0.14
    quo
    -0.13
    imes
    -0.13
    pecies
    -0.13
    uit
    -0.13
    ili
    -0.13
    ů
    -0.13
    ноÑģÑıÑĤ
    -0.12
    eman
    -0.12
     ettiÄŁi
    -0.12
    POSITIVE LOGITS
     sure
    0.64
     Sure
    0.42
    sure
    0.41
    Sure
    0.40
    SURE
    0.26
     sense
    0.23
     sû
    0.22
    -bel
    0.22
     sicher
    0.21
     it
    0.19
    Activations Density 0.079%

    No Known Activations