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

    Neuronpedia

    Jacobian LensNEW
    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. 12979
    Prev
    Next
    INDEX
    Explanations

    terms related to causation or actions leading to specific outcomes

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "caused" in various contexts

    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
    urch
    -0.63
     Zup
    -0.63
     draft
    -0.62
     nods
    -0.59
    ivot
    -0.59
    lder
    -0.57
    uggets
    -0.56
     emphasis
    -0.56
     finalists
    -0.56
     drafts
    -0.56
    POSITIVE LOGITS
     caused
    3.40
     triggered
    1.94
     provoked
    1.94
     inflicted
    1.83
     sparked
    1.81
     prompted
    1.68
     causing
    1.68
     resulted
    1.67
     exacerbated
    1.66
     causes
    1.66
    Activations Density 0.026%

    No Known Activations