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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24569
    Prev
    Next
    INDEX
    Explanations

    technical terms and concepts related to programming and localization

    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
    ëł¹
    -0.17
    DCALL
    -0.17
    SSIP
    -0.15
    ahoma
    -0.15
     Maul
    -0.14
    uria
    -0.14
    beck
    -0.14
    vier
    -0.14
    strom
    -0.14
    imir
    -0.14
    POSITIVE LOGITS
     SUS
    0.16
    รà¸ĵ
    0.15
    ent
    0.15
    erras
    0.15
    avan
    0.15
    (locale
    0.14
     tro
    0.14
     pathname
    0.14
    ิà¹Ĥ
    0.14
    urance
    0.14
    Activations Density 0.022%

    No Known Activations