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

    Neuronpedia

    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. 11-RES-JB
    11. 23262
    Prev
    Next
    INDEX
    Explanations

    The neuron primarily activates on the word "Those" before a list of descriptions or actions related to various subjects

    oai_token-act-pair · gpt-3.5-turbo

    the word "Those" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    PLA
    -0.80
    ori
    -0.74
    kamp
    -0.74
    iggurat
    -0.73
    obook
    -0.73
    abis
    -0.73
    efully
    -0.72
    ob
    -0.71
    andise
    -0.71
    Published
    -0.70
    POSITIVE LOGITS
     pesky
    1.24
     who
    1.22
    who
    1.06
     kinds
    0.97
     whom
    0.95
     sorts
    0.93
     wishing
    0.91
     caveats
    0.85
     whose
    0.85
     darn
    0.84
    Activations Density 0.068%

    No Known Activations