© 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. 63297
    Prev
    Next
    INDEX
    Explanations

    negative constructions and expressions of inability or 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
    ilim
    -0.14
     NÄĽm
    -0.12
    ---</
    -0.12
    виÑĤ
    -0.12
    andest
    -0.12
    /*č↵
    -0.12
    ذر
    -0.12
     rencont
    -0.11
    uien
    -0.11
    à¹ĩà¸ĩ
    -0.11
    POSITIVE LOGITS
    _exempt
    0.14
    Ī
    0.12
     ########.
    0.12
    Ļ
    0.11
    §
    0.11
    icky
    0.11
    _REF
    0.10
    egov
    0.10
    ±Ð¾ÑĤ
    0.10
    izen
    0.10
    Activations Density 0.872%

    No Known Activations