© 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. 54435
    Prev
    Next
    INDEX
    Explanations

    modal verbs indicating capability and planning

    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
    adil
    -0.18
    letics
    -0.16
    entine
    -0.15
    kat
    -0.15
    ado
    -0.15
    ozor
    -0.15
    aland
    -0.15
    kv
    -0.14
     als
    -0.14
    à¸ĩà¸Ħ
    -0.14
    POSITIVE LOGITS
     possibly
    0.20
     while
    0.19
     possible
    0.17
     given
    0.17
    given
    0.17
     POSS
    0.16
     within
    0.16
     manage
    0.16
     without
    0.16
     Compatibility
    0.15
    Activations Density 0.036%

    No Known Activations