© 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. 10-RES-JB
    11. 7779
    Prev
    Next
    INDEX
    Explanations

    words related to violence and criminal activities

    oai_token-act-pair · gpt-3.5-turbo

    discussions of violence against marginalized groups, particularly women

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    answer
    -0.85
    terday
    -0.81
    rar
    -0.79
    DragonMagazine
    -0.78
    xon
    -0.74
    ername
    -0.74
    fen
    -0.73
    uyomi
    -0.73
    forward
    -0.72
    senal
    -0.71
    POSITIVE LOGITS
     minors
    1.50
     humans
    1.36
     individuals
    1.35
     juveniles
    1.35
     adults
    1.35
     adolescents
    1.33
     females
    1.33
     persons
    1.33
     infants
    1.32
     minorities
    1.32
    Activations Density 0.361%

    No Known Activations