© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 15524
    Prev
    Next
    INDEX
    Explanations

    various forms of artistic expression and emotional depth

    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
    θι
    -0.14
    коп
    -0.14
     å¾Ĵ
    -0.14
    unca
    -0.13
    lesh
    -0.13
    _ASSUME
    -0.12
    WithType
    -0.12
    ,},↵
    -0.12
    WithString
    -0.12
    باÙĨ
    -0.12
    POSITIVE LOGITS
     and
    0.33
     &
    0.23
     и
    0.21
     vÃł
    0.21
    à¹ģละ
    0.20
     és
    0.20
    ãģ¨
    0.19
    åĴĮ
    0.18
     ÈĻi
    0.18
     factor
    0.18
    Activations Density 0.587%

    No Known Activations