© 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. 41787
    Prev
    Next
    INDEX
    Explanations

    references to locations and structures, particularly in the context of conflict

    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
    adro
    -0.14
    fdc
    -0.14
    ियत
    -0.13
    инки
    -0.13
     inval
    -0.13
    rana
    -0.13
    ابت
    -0.13
     interv
    -0.12
    ofilm
    -0.12
    importe
    -0.12
    POSITIVE LOGITS
     complex
    0.32
     located
    0.28
    complex
    0.26
     Complex
    0.26
    Complex
    0.24
     built
    0.23
     compound
    0.22
     facility
    0.21
     site
    0.21
    located
    0.21
    Activations Density 0.183%

    No Known Activations