© 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. 1-RES-JB
    11. 14478
    Prev
    Next
    INDEX
    Explanations

    words related to negative qualities or characteristics

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "ugly" or similar descriptors focusing on negative aesthetics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    owder
    -0.98
    akings
    -0.89
    ingham
    -0.86
    iasm
    -0.83
    ership
    -0.82
    aver
    -0.81
    eding
    -0.77
    thus
    -0.76
    idential
    -0.76
    funding
    -0.75
    POSITIVE LOGITS
     ugly
    1.40
     elephant
    0.87
     adolesc
    0.87
     spectacle
    0.78
     crap
    0.78
     fallout
    0.76
     nasty
    0.76
     losers
    0.76
     hairy
    0.75
     hideous
    0.75
    Activations Density 0.007%

    No Known Activations