© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 42020
    Prev
    Next
    INDEX
    Explanations

    future tense verbs, particularly forms of "will" and "be"

    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
    ojis
    -0.14
    ventus
    -0.14
    acco
    -0.14
    azu
    -0.14
    ''''
    -0.14
    bilt
    -0.13
    illions
    -0.13
    itol
    -0.13
    quets
    -0.13
    vÃŃc
    -0.13
    POSITIVE LOGITS
     be
    0.52
     soon
    0.29
    soon
    0.27
    l
    0.26
     Soon
    0.26
    iams
    0.24
    iam
    0.24
     Be
    0.22
    kommen
    0.21
     shortly
    0.21
    Activations Density 0.177%

    No Known Activations