© 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. 29508
    Prev
    Next
    INDEX
    Explanations

    negative expressions related to lack of ability or opportunity

    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
    Ø°Ùĩ
    -0.17
    ingles
    -0.15
    ilit
    -0.15
    gren
    -0.14
    erence
    -0.14
    ê°ķ
    -0.14
    yer
    -0.14
    tober
    -0.14
    _PUSHDATA
    -0.14
     defaultManager
    -0.13
    POSITIVE LOGITS
     until
    0.66
    until
    0.54
     Until
    0.53
    Until
    0.51
     till
    0.42
     hasta
    0.42
     até
    0.39
    _until
    0.38
    .until
    0.37
     jusqu
    0.36
    Activations Density 0.020%

    No Known Activations