© 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. 18184
    Prev
    Next
    INDEX
    Explanations

    quotations involving the word "could."

    oai_token-act-pair · gpt-3.5-turbo

    modal verbs indicating possibility or necessity

    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
    's
    -0.68
     started
    -0.68
     visible
    -0.68
     Association
    -0.66
    ,
    -0.66
     involved
    -0.65
     used
    -0.65
     featured
    -0.64
     rental
    -0.63
     created
    -0.63
    POSITIVE LOGITS
    would
    2.43
    could
    2.35
    should
    2.10
    does
    1.90
    was
    1.90
    didn
    1.83
    might
    1.82
    doesn
    1.78
    were
    1.71
    will
    1.64
    Activations Density 0.030%

    No Known Activations