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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9108
    Prev
    Next
    INDEX
    Explanations

    negations or forms of denial

    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
    equal
    -0.14
    ebek
    -0.14
    à¹Ĩ
    -0.14
    eval
    -0.13
     indeb
    -0.13
    eb
    -0.13
    ismet
    -0.13
    Âĸ
    -0.13
    ej
    -0.12
    ´
    -0.12
    POSITIVE LOGITS
    't
    0.41
    ’t
    0.39
    ning
    0.20
    _REF
    0.17
    uts
    0.17
    ´t
    0.17
    ’ta
    0.16
    ned
    0.16
    'T
    0.16
    ’hui
    0.15
    Activations Density 0.157%

    No Known Activations