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. 23644
    Prev
    Next
    INDEX
    Explanations

    punctuation marks that may signify key statements or conclusions

    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
    this
    -0.17
     therefore
    -0.17
    ãi
    -0.17
     thus
    -0.17
    andom
    -0.16
    thus
    -0.16
    olt
    -0.15
    ItemAt
    -0.15
     this
    -0.15
     ÙĪØ£ÙĨ
    -0.15
    POSITIVE LOGITS
     first
    0.28
     Firstly
    0.25
     firstly
    0.23
    first
    0.21
    é¦ĸ
    0.21
     because
    0.20
     imagine
    0.20
     primero
    0.20
    æ¯ķ
    0.19
    .first
    0.18
    Activations Density 0.038%

    No Known Activations