© 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. 9927
    Prev
    Next
    INDEX
    Explanations

    references to the word "how"

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "how."

    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
    stone
    -0.75
    onte
    -0.70
    soever
    -0.61
    onia
    -0.61
     ®
    -0.59
    itarian
    -0.58
    agues
    -0.57
    nos
    -0.56
    anto
    -0.55
    erenn
    -0.54
    POSITIVE LOGITS
     how
    2.91
    how
    2.15
     HOW
    1.93
     why
    1.86
     How
    1.74
    why
    1.61
    How
    1.59
     what
    1.53
     whether
    1.52
    HOW
    1.48
    Activations Density 0.076%

    No Known Activations