© 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. 54352
    Prev
    Next
    INDEX
    Explanations

    affirmations of collective achievement and responsibility

    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
     sympathy
    -0.17
     scape
    -0.15
     sympath
    -0.14
     Compass
    -0.14
    ÑĤон
    -0.14
    Proxy
    -0.14
    Transactional
    -0.13
     sympathetic
    -0.13
    ton
    -0.13
    Facade
    -0.13
    POSITIVE LOGITS
     pride
    0.90
     proud
    0.82
     Pride
    0.79
     Proud
    0.72
     proudly
    0.56
     brag
    0.55
     boast
    0.51
     boasting
    0.45
     boasted
    0.43
     boasts
    0.42
    Activations Density 0.038%

    No Known Activations