© 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. 23097
    Prev
    Next
    INDEX
    Explanations

    phrases indicating conditions and evaluations related to information or experiences

    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
     abych
    -0.18
     saya
    -0.17
     unseren
    -0.15
     meu
    -0.15
    .scalablytyped
    -0.15
     мой
    -0.14
     mio
    -0.14
     ÑįÑĤим
    -0.14
     minha
    -0.14
    à¸Ĥà¸Ńà¸ĩà¸ľ
    -0.14
    POSITIVE LOGITS
     his
    0.56
     himself
    0.54
     him
    0.48
     they
    0.47
     their
    0.44
    his
    0.43
     he
    0.41
     theirs
    0.40
    him
    0.39
     lui
    0.38
    Activations Density 0.029%

    No Known Activations