© 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. 30542
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "that said" used to introduce a contrasting point or clarification

    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
     thats
    -0.14
    uw
    -0.14
    lea
    -0.13
    .dateTime
    -0.13
    ','','
    -0.13
    ing
    -0.13
    _THAT
    -0.13
    éĤ£æł·
    -0.13
    uze
    -0.12
    anka
    -0.12
    POSITIVE LOGITS
    ernel
    0.18
    ripper
    0.17
    ebin
    0.15
    thic
    0.15
    ,↵↵↵↵
    0.15
    thest
    0.14
    resse
    0.14
    .DEFINE
    0.14
    :↵↵↵↵
    0.14
    åĩ¯
    0.14
    Activations Density 0.053%

    No Known Activations