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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 63436
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to published works and writings

    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
     dagen
    -0.12
    æ¼Ķ
    -0.12
     Paragraph
    -0.12
     effet
    -0.12
    ëħ
    -0.12
    аÑĢам
    -0.12
    /location
    -0.11
    579
    -0.11
     æ¼Ķ
    -0.11
    _Handler
    -0.11
    POSITIVE LOGITS
     volume
    0.28
     published
    0.27
     publication
    0.26
     volumes
    0.24
    volume
    0.23
    published
    0.23
     Volume
    0.23
    -volume
    0.23
     publish
    0.20
    Volume
    0.20
    Activations Density 0.071%

    No Known Activations