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. 23727
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing the act of making or ensuring something

    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
    adge
    -0.16
    egr
    -0.15
    _losses
    -0.14
    _BP
    -0.14
    wyn
    -0.14
    createClass
    -0.14
     infer
    -0.13
    to
    -0.13
    aga
    -0.13
     hát
    -0.13
    POSITIVE LOGITS
     sure
    0.35
     Sure
    0.27
    Sure
    0.24
    sure
    0.24
     clear
    0.23
     absolutely
    0.20
     crystal
    0.18
     Clear
    0.18
    -clear
    0.18
     perfectly
    0.18
    Activations Density 0.031%

    No Known Activations