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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 4-RES-JB
    6. 17739
    Prev
    Next
    INDEX
    Explanations

    phrases related to movies and entertainment industry, specifically focusing on appearances on the big screen

    oai_token-act-pair · gpt-3.5-turbo

    references to "screen" related to films or visual media

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     deleg
    -0.69
    ipop
    -0.68
     Esk
    -0.67
    onomic
    -0.65
     Upton
    -0.63
     Harriet
    -0.63
     grieving
    -0.63
    ton
    -0.61
     accrued
    -0.60
     horrend
    -0.60
    POSITIVE LOGITS
    plays
    1.28
    writer
    1.13
    writers
    1.11
    screen
    1.01
     TVs
    0.96
    writing
    0.95
     screens
    0.89
    Screen
    0.86
     Screen
    0.81
    gallery
    0.81
    Activations Density 0.013%

    No Known Activations