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. 31407
    Prev
    Next
    INDEX
    Explanations

    historical references and significant events

    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
     Millennium
    -0.20
    ï¼Īå¹³æĪIJ
    -0.19
    199
    -0.17
    ushima
    -0.16
     kak
    -0.15
     Interactive
    -0.15
    200
    -0.15
     Lexus
    -0.14
    Interactive
    -0.14
    Û±Û¹Û¹
    -0.14
    POSITIVE LOGITS
    196
    0.78
    195
    0.57
    Û±Û¹Û¶
    0.47
    197
    0.39
     Nixon
    0.38
     Eisenhower
    0.33
     Vietnam
    0.33
     Beatles
    0.32
     hipp
    0.32
    Û±Û¹Ûµ
    0.32
    Activations Density 0.022%

    No Known Activations