© 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 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. 43359
    Prev
    Next
    INDEX
    Explanations

    expressions of positive sentiment or good news

    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
     seperate
    -0.22
    Amt
    -0.20
     seper
    -0.20
     amt
    -0.19
     thereby
    -0.16
    .utility
    -0.15
    ubl
    -0.14
    _via
    -0.14
    .construct
    -0.14
    okay
    -0.14
    POSITIVE LOGITS
     splitted
    0.24
     confort
    0.24
    wich
    0.20
    _infos
    0.19
     serie
    0.18
     aprox
    0.17
     mant
    0.17
    nal
    0.17
     negoci
    0.17
    ´s
    0.17
    Activations Density 0.074%

    No Known Activations