© 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
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 41614
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate lessons learned or insights gained

    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
    angen
    -0.15
    otive
    -0.15
    atus
    -0.14
    ubar
    -0.14
    uet
    -0.13
    .robot
    -0.13
    ismatch
    -0.13
    Ïģιά
    -0.13
    åįĴ
    -0.13
    enheim
    -0.12
    POSITIVE LOGITS
     lessons
    1.00
     lesson
    1.00
     Lesson
    0.88
     Lessons
    0.85
    Lesson
    0.78
    lessons
    0.77
    lesson
    0.76
     ÑĥÑĢок
    0.57
     learn
    0.53
    .lesson
    0.50
    Activations Density 0.029%

    No Known Activations