© 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. 41033
    Prev
    Next
    INDEX
    Explanations

    instances of the word "some" and related terms indicating ambiguity or inconsistency

    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
    اÙĨÙĩ
    -0.18
    las
    -0.17
     somehow
    -0.17
     respectively
    -0.16
    λλι
    -0.16
    åIJĦç§į
    -0.15
     someone
    -0.15
    enate
    -0.15
     something
    -0.15
    ä½ķãģĭ
    -0.14
    POSITIVE LOGITS
    ones
    0.39
     of
    0.36
    /all
    0.35
    place
    0.34
    hw
    0.27
    ONE
    0.23
     degree
    0.23
    -times
    0.23
    ONES
    0.21
    obody
    0.21
    Activations Density 0.135%

    No Known Activations