© 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. 2431
    Prev
    Next
    INDEX
    Explanations

    sentences describing violent events or crimes

    oai_token-act-pair · gpt-3.5-turbo

    incidents related to crime or violence

    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
     advantageous
    -0.75
     condensed
    -0.73
     freely
    -0.73
     indispensable
    -0.72
     excel
    -0.71
     manageable
    -0.69
    ":"/
    -0.69
     optimized
    -0.68
     useful
    -0.67
     intellig
    -0.67
    POSITIVE LOGITS
    Elsewhere
    1.03
     Another
    1.03
    Two
    1.00
    Another
    0.98
     Earlier
    0.95
    two
    0.92
    Similarly
    0.88
     Two
    0.87
    Three
    0.87
     Similarly
    0.86
    Activations Density 0.377%

    No Known Activations