© 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 EvalsExportsAPI 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. 14831
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "piece together" or similar formulations indicating assembling or creating something from various parts

    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
    fty
    -0.15
    bad
    -0.14
    vil
    -0.14
     Loot
    -0.14
    ars
    -0.14
    robat
    -0.14
    edir
    -0.14
     teeth
    -0.13
     Buch
    -0.13
    ucz
    -0.13
    POSITIVE LOGITS
     together
    0.41
     Together
    0.32
    Together
    0.30
    gether
    0.28
     zusammen
    0.22
     вмеÑģÑĤе
    0.20
    ä¸Ģèµ·
    0.19
     sammen
    0.18
    (patch
    0.17
     patch
    0.17
    Activations Density 0.007%

    No Known Activations