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. 0-RES-JB
    6. 19542
    Prev
    Next
    INDEX
    Explanations

    references to different parts or aspects of a house, such as specific areas or activities related to a house

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of "house."

    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
    imedia
    -0.68
     Roberts
    -0.66
    rint
    -0.66
    idem
    -0.65
    ova
    -0.64
    Downloadha
    -0.63
    asive
    -0.63
    iliated
    -0.62
     Armstrong
    -0.61
    milo
    -0.61
    POSITIVE LOGITS
     house
    3.63
     houses
    2.67
    house
    2.22
     HOUSE
    2.21
    House
    1.98
     House
    1.97
     mansion
    1.94
     Houses
    1.93
    houses
    1.69
     apartment
    1.55
    Activations Density 0.016%

    No Known Activations