© 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. 23819
    Prev
    Next
    INDEX
    Explanations

    references to television events and specials

    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
    ãĥĩãĤ£ãĤ¢
    -0.16
     Rosenberg
    -0.16
    елÑĮзÑı
    -0.15
    iez
    -0.14
     circ
    -0.14
    зм
    -0.13
     Wass
    -0.13
    á»ijt
    -0.13
    iais
    -0.13
    GIN
    -0.13
    POSITIVE LOGITS
     specials
    0.43
     special
    0.42
    special
    0.37
     Special
    0.32
    Special
    0.30
     segment
    0.30
    -special
    0.28
    PECIAL
    0.27
     SPECIAL
    0.27
     segments
    0.25
    Activations Density 0.049%

    No Known Activations