© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 9112
    Prev
    Next
    INDEX
    Explanations

    expressions of amazement or surprise

    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
    shine
    -0.14
    ÑģÑĤÑĢой
    -0.14
     Ïĥελ
    -0.14
     wir
    -0.14
    夢
    -0.14
    olec
    -0.14
     Sad
    -0.13
    ickness
    -0.13
    utomation
    -0.13
     contrad
    -0.13
    POSITIVE LOGITS
    htaking
    0.20
    ingly
    0.19
     Alter
    0.19
    icrous
    0.18
    -def
    0.18
    slaught
    0.17
    -threatening
    0.17
    -num
    0.17
    attering
    0.17
    -sh
    0.17
    Activations Density 0.044%

    No Known Activations