Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 24222
    Prev
    Next
    INDEX
    Explanations

    words related to conflict or aggression, specifically focusing on the term "tantrum."

    oai_token-act-pair · gpt-3.5-turbo

    words related to tantrums and emotional outbursts

    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
    Not in Any Lists

    No Comments

    Negative Logits
     surn
    -0.75
     recru
    -0.72
    cest
    -0.70
    IRC
    -0.69
     guessed
    -0.67
     recruit
    -0.67
    isen
    -0.66
    chart
    -0.66
    Mobile
    -0.65
     rename
    -0.65
    POSITIVE LOGITS
     tant
    1.81
    andal
    1.43
     withstand
    1.41
     Tant
    1.33
     reflective
    1.25
     mas
    1.23
    aband
    1.04
    andals
    1.02
     Dangerous
    0.97
     Moment
    0.88
    Activations Density 0.067%

    No Known Activations