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. 3-RES-JB
    6. 19053
    Prev
    Next
    INDEX
    Explanations

    specific words related to facilities or buildings

    oai_token-act-pair · gpt-3.5-turbo

    references to facilities or structures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    plets
    -0.69
    poll
    -0.68
    ohan
    -0.66
    jin
    -0.66
    thin
    -0.63
     vengeance
    -0.63
    haps
    -0.62
    quel
    -0.62
    ju
    -0.62
     paras
    -0.62
    POSITIVE LOGITS
     facility
    0.98
     facilities
    0.91
     Facility
    0.80
    ility
    0.73
     premises
    0.72
     Facilities
    0.71
    imore
    0.69
    ounter
    0.68
     utilization
    0.67
     renovations
    0.67
    Activations Density 0.014%

    No Known Activations