© 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. 24005
    Prev
    Next
    INDEX
    Explanations

    terms related to smoothness or seamlessness in context

    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
    izzato
    -0.14
    बल
    -0.14
    ongan
    -0.14
     scope
    -0.14
    ète
    -0.14
    .createClass
    -0.14
     seedu
    -0.14
    ETA
    -0.13
    anth
    -0.13
    èIJ
    -0.13
    POSITIVE LOGITS
     transition
    0.39
     transitions
    0.37
    transition
    0.33
     Transition
    0.33
    Transition
    0.29
    -transition
    0.29
     transitioning
    0.28
    -flow
    0.25
    .transition
    0.24
    _transition
    0.22
    Activations Density 0.016%

    No Known Activations