© 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. 29035
    Prev
    Next
    INDEX
    Explanations

    references to inflammatory speech and rhetoric that targets groups or individuals

    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
    quential
    -0.15
    atto
    -0.15
    ismet
    -0.14
     punishing
    -0.14
    aghetti
    -0.14
     rede
    -0.14
    ãĥĩãĥ«
    -0.14
    ifix
    -0.14
    isz
    -0.14
    amel
    -0.13
    POSITIVE LOGITS
     inst
    0.43
     inc
    0.36
    Inst
    0.30
     Inst
    0.30
    inst
    0.29
     fuel
    0.29
     infl
    0.28
     stir
    0.27
    çħ
    0.26
     stirring
    0.25
    Activations Density 0.061%

    No Known Activations