© 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. 0-RES-JB
    11. 7792
    Prev
    Next
    INDEX
    Explanations

    phrases related to love and kindness

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "loving" and related concepts of love and care

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •layer 0 - love

    No Comments

    Negative Logits
     Auditor
    -0.68
    inar
    -0.67
    =[
    -0.64
     Audit
    -0.64
     Procedures
    -0.64
    cture
    -0.63
    rium
    -0.62
    UMP
    -0.61
    acket
    -0.61
    UA
    -0.61
    POSITIVE LOGITS
     loving
    3.76
     Loving
    2.27
    loving
    2.03
     caring
    1.72
    love
    1.67
     love
    1.67
     hating
    1.62
     loved
    1.58
     LOVE
    1.48
     liking
    1.48
    Activations Density 0.018%

    No Known Activations