© 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. 8991
    Prev
    Next
    INDEX
    Explanations

    expressions that discuss the nature of religion and violence

    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
     tiá»ĩn
    -0.14
    aille
    -0.14
     Convenient
    -0.14
    çĶ²
    -0.14
    над
    -0.13
    Opaque
    -0.13
     çĭ
    -0.13
    comfort
    -0.13
    oldt
    -0.13
     conspir
    -0.13
    POSITIVE LOGITS
     peaceful
    0.65
     peace
    0.60
    peace
    0.55
     Peace
    0.54
    Peace
    0.54
     peacefully
    0.50
     pac
    0.48
     gentle
    0.38
    pac
    0.36
     Pac
    0.35
    Activations Density 0.042%

    No Known Activations