© 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. 15658
    Prev
    Next
    INDEX
    Explanations

    ratings and assessments of health or condition quality

    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
     ÐļÑĢÑĸм
    -0.15
    ÏĢοÏį
    -0.14
    amm
    -0.14
     younger
    -0.14
    czy
    -0.14
    illard
    -0.13
    éļĨ
    -0.13
     Maher
    -0.13
    éĽĨä¸Ń
    -0.13
     cazzo
    -0.13
    POSITIVE LOGITS
     medium
    0.45
     moderate
    0.41
    medium
    0.41
     Medium
    0.40
     Moderate
    0.38
    Medium
    0.37
    Moder
    0.34
    edium
    0.31
     intermediate
    0.31
    -medium
    0.31
    Activations Density 0.082%

    No Known Activations