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

    Neuronpedia

    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. 24526
    Prev
    Next
    INDEX
    Explanations

    phrases suggesting uncertainty or conditional evaluations

    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
    ohl
    -0.16
    pei
    -0.15
    onica
    -0.14
    bookmark
    -0.14
    elman
    -0.14
    WithTitle
    -0.13
     Missing
    -0.13
    prites
    -0.13
     missing
    -0.13
    ;element
    -0.13
    POSITIVE LOGITS
     better
    0.84
    better
    0.74
     best
    0.73
     Better
    0.68
    Better
    0.66
     mejor
    0.65
     melhor
    0.58
    best
    0.58
     besser
    0.55
     mieux
    0.55
    Activations Density 0.023%

    No Known Activations