© 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. 54280
    Prev
    Next
    INDEX
    Explanations

    expressions of exceptional quality or admiration

    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
    одÑĥ
    -0.15
    ération
    -0.15
    iga
    -0.14
    iston
    -0.14
    quia
    -0.14
    odoxy
    -0.14
     wise
    -0.14
    illation
    -0.13
    verity
    -0.13
    istem
    -0.13
    POSITIVE LOGITS
    idable
    0.23
    ificent
    0.22
    astic
    0.22
    acular
    0.20
    orable
    0.20
    ocious
    0.18
    cellent
    0.18
    ificant
    0.18
    orous
    0.18
    eworthy
    0.17
    Activations Density 0.075%

    No Known Activations