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. 31330
    Prev
    Next
    INDEX
    Explanations

    references to specific years

    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
     
    -0.23
    ergy
    -0.17
    336
    -0.16
    iface
    -0.16
    ource
    -0.15
     Stall
    -0.14
     enc
    -0.14
     wed
    -0.14
    43
    -0.14
    ero
    -0.14
    POSITIVE LOGITS
    ulg
    0.17
    _Anim
    0.16
    麼
    0.16
    uzzi
    0.16
    etr
    0.15
    ë²Į
    0.14
    INGLE
    0.14
    fur
    0.14
    odiac
    0.14
    reeNode
    0.14
    Activations Density 0.027%

    No Known Activations