© 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. 0-RES-JB
    11. 1478
    Prev
    Next
    INDEX
    Explanations

    verbs indicating an action or decision

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "take"

    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
    amer
    -0.73
     Development
    -0.61
    Development
    -0.60
    ector
    -0.60
    herent
    -0.60
    gat
    -0.59
    ames
    -0.59
    inguished
    -0.59
    hid
    -0.59
    isp
    -0.59
    POSITIVE LOGITS
     take
    2.92
    take
    2.27
     Take
    1.98
    Take
    1.93
     takes
    1.74
     took
    1.45
     seize
    1.40
     retake
    1.39
     taking
    1.37
     follow
    1.32
    Activations Density 0.054%

    No Known Activations