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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 23277
    Prev
    Next
    INDEX
    Explanations

    references to historical music figures and their works

    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
     Victorian
    -0.23
    188
    -0.23
    nict
    -0.21
    187
    -0.21
    185
    -0.20
    189
    -0.19
     Alfred
    -0.19
     medieval
    -0.19
     Medieval
    -0.18
    186
    -0.17
    POSITIVE LOGITS
    176
    0.51
    174
    0.51
    175
    0.47
    177
    0.43
    173
    0.42
    178
    0.38
    172
    0.37
     Enlightenment
    0.33
    171
    0.33
    179
    0.31
    Activations Density 0.072%

    No Known Activations