© 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. 29267
    Prev
    Next
    INDEX
    Explanations

    phrases indicating limitations or constraints in various contexts

    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
    ¶Į
    -0.16
    ç§
    -0.15
    ắt
    -0.15
    ัà¸Ļà¸ĺ
    -0.14
    ãĥ³ãĥĶ
    -0.13
    ÌĨ
    -0.13
    readcr
    -0.13
    urai
    -0.13
     sovere
    -0.12
    ÏĥÏį
    -0.12
    POSITIVE LOGITS
     importance
    0.38
     extent
    0.28
     Importance
    0.25
     dangers
    0.25
     role
    0.24
     lengths
    0.23
     impact
    0.23
     difference
    0.23
     danger
    0.21
     degree
    0.21
    Activations Density 0.289%

    No Known Activations