© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 25045
    Prev
    Next
    INDEX
    Explanations

    references to publishing information, particularly related to works and publications

    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
    akes
    -0.14
    еÑĢеж
    -0.14
     activated
    -0.14
    alous
    -0.14
     organic
    -0.14
     passive
    -0.13
    istani
    -0.13
    ãģ¨ãģĦ
    -0.13
    λει
    -0.13
    Hosting
    -0.13
    POSITIVE LOGITS
    197
    0.29
    196
    0.27
     Doub
    0.24
    198
    0.22
    195
    0.20
    Û±Û¹Û·
    0.18
     Random
    0.18
    éº
    0.17
    Random
    0.17
     Harper
    0.16
    Activations Density 0.040%

    No Known Activations