© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 12913
    Prev
    Next
    INDEX
    Explanations

    predictive language, specifically focusing on future actions and expectations

    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
    \Block
    -0.15
    rup
    -0.14
    raud
    -0.14
     Hin
    -0.14
    SetColor
    -0.13
    uya
    -0.13
    ÙĦسÙĦ
    -0.13
    QT
    -0.13
    ilis
    -0.13
    quel
    -0.13
    POSITIVE LOGITS
     tests
    0.60
     test
    0.59
     testing
    0.58
     tested
    0.52
    -testing
    0.50
     challenge
    0.50
     Testing
    0.48
     tester
    0.48
     Tests
    0.46
    æµĭè¯ķ
    0.45
    Activations Density 0.051%

    No Known Activations