© 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. 9134
    Prev
    Next
    INDEX
    Explanations

    expressions of pacing and structure in narratives

    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
    agers
    -0.15
     Stout
    -0.14
    RuleContext
    -0.14
    shiv
    -0.14
     prv
    -0.13
     McCart
    -0.13
    ffen
    -0.13
    abox
    -0.13
    imary
    -0.13
    ÅĽnie
    -0.13
    POSITIVE LOGITS
     pace
    0.84
     Pace
    0.69
     tempo
    0.63
    pace
    0.60
     pacing
    0.53
     paced
    0.52
    -paced
    0.52
     rhythm
    0.51
     Tempo
    0.51
    paced
    0.45
    Activations Density 0.037%

    No Known Activations