© 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. 22193
    Prev
    Next
    INDEX
    Explanations

    phrases related to estimation and evaluation of future impacts, particularly in a scientific or ecological context

    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
    assel
    -0.16
    ForObject
    -0.14
    .nt
    -0.13
    bron
    -0.13
     stained
    -0.13
    abbage
    -0.13
    elves
    -0.13
    oreach
    -0.13
    ergus
    -0.13
    olated
    -0.13
    POSITIVE LOGITS
     under
    0.76
     Under
    0.68
    under
    0.68
    Under
    0.68
     UNDER
    0.64
    -under
    0.60
    UNDER
    0.56
     underst
    0.55
     unders
    0.54
    _under
    0.54
    Activations Density 0.060%

    No Known Activations