© 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. 30178
    Prev
    Next
    INDEX
    Explanations

    adjectives and phrases that describe the quality or nature of objects or concepts

    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
     itself
    -0.20
    å®ĥ
    -0.17
    ä¹ĭä¸Ģ
    -0.16
     Taxes
    -0.15
    zer
    -0.14
    /Graphics
    -0.14
    uw
    -0.14
    Ø°ÙĬ
    -0.14
    ologue
    -0.13
     Bearings
    -0.13
    POSITIVE LOGITS
     themselves
    0.42
     ones
    0.28
     entities
    0.26
     their
    0.22
    Their
    0.22
     enough
    0.22
     Their
    0.21
    their
    0.21
     Ones
    0.20
     affairs
    0.20
    Activations Density 0.377%

    No Known Activations