Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 54685
    Prev
    Next
    INDEX
    Explanations

    mentions of structural instability and danger in buildings

    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
    ãĥ³ãĥĩ
    -0.17
    dna
    -0.16
    ContentSize
    -0.16
    zzo
    -0.16
    oyer
    -0.16
     hã
    -0.15
    oria
    -0.15
    strup
    -0.14
    stial
    -0.14
    รม
    -0.14
    POSITIVE LOGITS
     unstable
    0.51
     instability
    0.49
     stability
    0.48
     collapse
    0.45
     Stability
    0.42
     stable
    0.40
     Collapse
    0.39
     collapses
    0.38
     collapsing
    0.37
    stable
    0.37
    Activations Density 0.084%

    No Known Activations