© 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. 32929
    Prev
    Next
    INDEX
    Explanations

    phrases indicating unprecedented events or accomplishments

    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
    ãĥŃãĥ³
    -0.17
     Actual
    -0.16
    ameleon
    -0.15
    bane
    -0.15
    à¸Ļà¸ķ
    -0.15
    erken
    -0.14
    neys
    -0.14
    emek
    -0.14
    /Runtime
    -0.14
    až
    -0.14
    POSITIVE LOGITS
     seen
    0.40
    seen
    0.38
     Seen
    0.33
    Seen
    0.33
     anything
    0.30
     previously
    0.29
     before
    0.28
     previous
    0.27
    _seen
    0.26
    anything
    0.26
    Activations Density 0.043%

    No Known Activations