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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 41725
    Prev
    Next
    INDEX
    Explanations

    references to groups or communities of people

    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
    illo
    -0.15
    ãģĿãĤĮãģ¯
    -0.14
    oref
    -0.14
    101
    -0.14
    ork
    -0.14
    atur
    -0.13
    ater
    -0.13
    ure
    -0.13
    104
    -0.13
    .bluetooth
    -0.13
    POSITIVE LOGITS
     at
    0.64
     chez
    0.49
     tại
    0.48
     bei
    0.39
    	at
    0.32
     beim
    0.30
     ợ
    0.30
    èĩ³å°ij
    0.30
     hos
    0.28
    at
    0.27
    Activations Density 0.047%

    No Known Activations