Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 23136
    Prev
    Next
    INDEX
    Explanations

    words related to changes or transitions in status or condition

    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
    à¹Ģà¸Ļ
    -0.16
    ajo
    -0.14
    wolf
    -0.14
    Fullscreen
    -0.14
    à¸Ĺำ
    -0.14
    prak
    -0.13
    iggs
    -0.13
    ONY
    -0.13
    ợ
    -0.13
    éli
    -0.13
    POSITIVE LOGITS
     for
    0.23
     dla
    0.20
     FAST
    0.17
     elsewhere
    0.17
     Favor
    0.17
     unfavor
    0.17
     differently
    0.17
    for
    0.17
    ÙĦÙĦ
    0.16
     fast
    0.16
    Activations Density 0.066%

    No Known Activations