© 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. 32907
    Prev
    Next
    INDEX
    Explanations

    phrases related to government actions and legal constraints

    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
    edes
    -0.15
    achsen
    -0.15
    erness
    -0.14
     облиÑĩ
    -0.14
    aska
    -0.14
    ủng
    -0.14
     Wanna
    -0.14
     Kurum
    -0.14
     debunk
    -0.14
    egov
    -0.14
    POSITIVE LOGITS
     simply
    0.28
     just
    0.25
     einfach
    0.24
     simplement
    0.20
     caval
    0.20
    just
    0.20
     suddenly
    0.19
     пÑĢоÑģÑĤо
    0.17
     Simply
    0.17
     deciding
    0.17
    Activations Density 0.086%

    No Known Activations