© 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. 15498
    Prev
    Next
    INDEX
    Explanations

    Expressions that indicate contrast or exceptions

    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
    ence
    -0.16
    dal
    -0.15
     Contin
    -0.15
    lier
    -0.14
     accordingly
    -0.14
    scribe
    -0.14
    ãģ³
    -0.13
     discre
    -0.13
    ì§ģ
    -0.13
    æĮĻ
    -0.13
    POSITIVE LOGITS
     also
    0.30
    also
    0.29
    Also
    0.28
     Also
    0.26
     necessary
    0.26
     Necessary
    0.26
    necessary
    0.24
     necessity
    0.22
     también
    0.21
     ALSO
    0.20
    Activations Density 0.059%

    No Known Activations