© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 2-RES-JB
    11. 15990
    Prev
    Next
    INDEX
    Explanations

    references to community locations or groups within contexts such as neighborhoods, towns, or regions

    oai_token-act-pair · gpt-3.5-turbo

    mentions of communities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    iasis
    -0.83
    cer
    -0.77
    nit
    -0.71
    Tro
    -0.69
    ifer
    -0.66
    eton
    -0.65
    tein
    -0.64
    BN
    -0.63
    Men
    -0.62
     Verb
    -0.61
    POSITIVE LOGITS
     councils
    1.08
    chool
    1.04
     communities
    1.01
     Communities
    0.84
    mith
    0.79
    wana
    0.79
    folk
    0.76
     neighbourhoods
    0.73
     subreddits
    0.73
    ensical
    0.73
    Activations Density 0.011%

    No Known Activations