© 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 EvalsExportsinterp-engineNEWAPI 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. 8-RES-JB
    11. 12531
    Prev
    Next
    INDEX
    Explanations

    phrases indicating skepticism or doubt

    oai_token-act-pair · gpt-3.5-turbo

    statements related to conditions or situations that are ongoing or relevant

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ulus
    -0.87
    uld
    -0.82
    erva
    -0.79
    alus
    -0.78
    culosis
    -0.74
    oting
    -0.73
    wi
    -0.71
    gow
    -0.71
    MQ
    -0.69
    ulla
    -0.68
    POSITIVE LOGITS
     intertwined
    0.94
     legion
    0.94
     amplified
    0.91
     plentiful
    0.91
     irrelevant
    0.90
     interchangeable
    0.90
     inconsistent
    0.87
     reversed
    0.86
     identical
    0.84
     unavoidable
    0.84
    Activations Density 0.298%

    No Known Activations