© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 22522
    Prev
    Next
    INDEX
    Explanations

    negative statements or controversies related to public figures

    oai_token-act-pair · gpt-3.5-turbo

    derogatory terms and phrases related to social issues and public figures

    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
     prepar
    -0.70
     Incre
    -0.63
    igree
    -0.62
    vantage
    -0.61
    eworks
    -0.59
    yrinth
    -0.59
    rieve
    -0.59
    accompan
    -0.59
    ilitation
    -0.58
     synerg
    -0.58
    POSITIVE LOGITS
     sexist
    1.26
     racist
    1.19
     homophobic
    1.18
     misogyny
    1.18
     misogyn
    1.17
     racists
    1.14
     sexism
    1.10
     feminists
    1.10
     homophobia
    1.10
     slurs
    1.09
    Activations Density 1.010%

    No Known Activations