© 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. 8529
    Prev
    Next
    INDEX
    Explanations

    references to characters and noteworthy elements in comic book narratives

    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
    appers
    -0.17
    æķħ
    -0.17
    ãĥ³ãĤ°ãĥ«
    -0.16
    lex
    -0.16
    quest
    -0.16
     æķħ
    -0.15
    afone
    -0.15
    esty
    -0.14
    à¹īาà¸Ĭ
    -0.14
    umbo
    -0.14
    POSITIVE LOGITS
    odore
    0.23
    odor
    0.20
    oden
    0.19
     Mask
    0.19
     Sand
    0.18
    odos
    0.16
     titular
    0.16
     Execution
    0.16
    od
    0.16
     original
    0.15
    Activations Density 0.088%

    No Known Activations