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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 0-RES-JB
    6. 17758
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to locations and titles

    oai_token-act-pair · gpt-3.5-turbo

    names of places and people, particularly related to Tennessee and Massachusetts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    plot
    -0.72
     foundations
    -0.67
     Bog
    -0.67
    hello
    -0.66
     NAD
    -0.65
     inert
    -0.65
    sych
    -0.65
    boost
    -0.64
    dark
    -0.64
     dreaming
    -0.62
    POSITIVE LOGITS
    essee
    4.76
    idents
    1.59
    ippi
    1.39
    etts
    1.28
    achusetts
    1.27
    omever
    1.25
    ittee
    1.21
    icut
    1.10
    orneys
    1.06
    aments
    0.95
    Activations Density 0.029%

    No Known Activations