© 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. 29496
    Prev
    Next
    INDEX
    Explanations

    instances of the word "quell" and its variations, indicating a focus on themes of suppression or control

    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
    μÎŃ
    -0.16
    raz
    -0.15
    aliz
    -0.14
    obot
    -0.14
    argin
    -0.14
    stry
    -0.14
    åı¤
    -0.13
    .mixer
    -0.13
    istring
    -0.13
    igh
    -0.13
    POSITIVE LOGITS
    orum
    0.18
    atern
    0.17
    ench
    0.17
    attro
    0.16
    eti
    0.16
    ivalent
    0.16
    adr
    0.15
    aggi
    0.15
    ero
    0.15
    rels
    0.15
    Activations Density 0.011%

    No Known Activations