© 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. 7-RES-JB
    11. 2962
    Prev
    Next
    INDEX
    Explanations

    negative words or phrases, focusing on the concept of meanness

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with "mean" or the concept of being unkind or harsh

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    conservancy
    -0.89
    ADRA
    -0.86
     accomp
    -0.81
    @#&
    -0.75
     mathemat
    -0.75
    ASED
    -0.75
    FINE
    -0.75
    UNCH
    -0.71
    ournal
    -0.69
     DRAG
    -0.67
    POSITIVE LOGITS
    est
    1.28
    ness
    1.27
    nesses
    1.19
    ings
    1.06
    ening
    1.05
     spirited
    1.04
    ingly
    1.00
    eness
    1.00
    liness
    1.00
    itude
    0.99
    Activations Density 0.053%

    No Known Activations