© 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. 3-RES-JB
    11. 158
    Prev
    Next
    INDEX
    Explanations

    questions or interrogative sentences

    oai_token-act-pair · gpt-3.5-turbo

    questions and inquiries, particularly those starting with "what," "where," "why," and "does."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     irony
    -0.64
    icides
    -0.63
    PF
    -0.63
     Literature
    -0.63
    ORPG
    -0.62
     supra
    -0.62
     PRESS
    -0.61
    milo
    -0.61
    ¶æ
    -0.60
    ibrary
    -0.60
    POSITIVE LOGITS
    inki
    0.80
    enos
    0.76
    ilib
    0.71
    hetic
    0.71
    ertodd
    0.68
    hel
    0.67
    het
    0.67
    hesive
    0.67
     Bakr
    0.66
     unavoid
    0.66
    Activations Density 0.350%

    No Known Activations