© 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. 23948
    Prev
    Next
    INDEX
    Explanations

    references to speculation and educated guesses

    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
    ahl
    -0.15
    urai
    -0.14
    yro
    -0.14
    itä
    -0.14
    ifier
    -0.14
    maal
    -0.13
    à¹ģล
    -0.13
    orton
    -0.13
    ombies
    -0.13
    è¦ĸ
    -0.13
    POSITIVE LOGITS
     guess
    0.71
     Guess
    0.62
    guess
    0.61
     guesses
    0.59
    Guess
    0.57
     guessing
    0.54
    _guess
    0.51
     guessed
    0.48
     educated
    0.47
    çĮľ
    0.42
    Activations Density 0.047%

    No Known Activations