© 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. 15739
    Prev
    Next
    INDEX
    Explanations

    qualities that contribute to the performance of a tablet

    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
     Gregory
    -0.16
    alar
    -0.15
    uktur
    -0.15
     safely
    -0.14
    é¡į
    -0.14
    æ¿ĥ
    -0.14
    éĩįè¤ĩ
    -0.14
     Axel
    -0.14
     çĽ
    -0.13
    гаÑĢ
    -0.13
    POSITIVE LOGITS
     smooth
    0.47
     Smooth
    0.39
    smooth
    0.38
    Smooth
    0.37
     fluid
    0.34
     smoother
    0.33
    fluid
    0.29
     smoothing
    0.28
     Fluid
    0.28
    Fluid
    0.28
    Activations Density 0.112%

    No Known Activations