© 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. 54893
    Prev
    Next
    INDEX
    Explanations

    expressions of positive achievements and noteworthy efforts

    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
    due
    -0.17
     Meaning
    -0.16
    meaning
    -0.16
     meaning
    -0.14
    еко
    -0.14
    /live
    -0.14
    RestController
    -0.13
    ilarity
    -0.13
     ÑĤобÑĤо
    -0.13
    iosa
    -0.13
    POSITIVE LOGITS
     especially
    0.45
    especially
    0.36
     particularly
    0.31
     but
    0.30
     especialmente
    0.29
     Especially
    0.28
    å°¤
    0.27
    particularly
    0.25
     оÑģобенно
    0.23
     BUT
    0.22
    Activations Density 0.149%

    No Known Activations