© 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. 22199
    Prev
    Next
    INDEX
    Explanations

    descriptions of people using positive or complimentary adjectives and nouns

    oai_token-act-pair · gpt-3.5-turbo

    positive descriptors of people, especially focusing on characteristics of men and women

    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
     VIDEOS
    -0.69
    gur
    -0.68
     ratification
    -0.63
    IDs
    -0.62
    uld
    -0.61
     virginity
    -0.60
    abies
    -0.59
    ormons
    -0.57
    iners
    -0.57
    skirts
    -0.57
    POSITIVE LOGITS
     who
    1.26
     whose
    1.20
     whom
    1.12
    who
    1.03
     capable
    1.00
    whose
    0.97
    nered
    0.97
     deserving
    0.95
     prone
    0.90
     devoid
    0.90
    Activations Density 0.196%

    No Known Activations