© 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. 14818
    Prev
    Next
    INDEX
    Explanations

    instances of speculation and questions regarding possibilities and uncertainties

    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
    oku
    -0.15
    Č↵
    -0.15
     ë³´ë©´
    -0.14
    ierarchy
    -0.14
    aggio
    -0.14
    á¿
    -0.14
    ybrid
    -0.14
     Pruitt
    -0.14
     RootState
    -0.14
    avra
    -0.14
    POSITIVE LOGITS
     Spec
    0.59
     spec
    0.58
     speculation
    0.58
    spec
    0.53
    Spec
    0.52
     speculative
    0.48
    -spec
    0.48
     speculate
    0.47
    _spec
    0.41
     speculated
    0.41
    Activations Density 0.049%

    No Known Activations