© 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. 14442
    Prev
    Next
    INDEX
    Explanations

    expressions related to physical contact with the ground or surface

    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
    ë°±
    -0.14
     submar
    -0.14
    _SEG
    -0.14
    ht
    -0.14
    anik
    -0.13
     misunder
    -0.13
    rame
    -0.13
    ifen
    -0.13
    moth
    -0.13
    ủ
    -0.13
    POSITIVE LOGITS
     ground
    0.88
    ground
    0.67
    Ground
    0.63
     Ground
    0.62
     floor
    0.58
    -ground
    0.58
    GROUND
    0.53
    _ground
    0.52
     Floor
    0.47
    floor
    0.45
    Activations Density 0.047%

    No Known Activations