© 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 EvalsExportsinterp-engineNEWAPI 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. 10-RES-JB
    11. 18901
    Prev
    Next
    INDEX
    Explanations

    phrases where the word "not" is followed by a numerical score denoting the strength of negation

    oai_token-act-pair · gpt-3.5-turbo

    the word "not" and its variations in context to negation or denial

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    imei
    -0.70
    nect
    -0.69
    adra
    -0.67
    kees
    -0.62
    æĸ¹
    -0.61
    eger
    -0.61
     Oath
    -0.60
    âĹ¼
    -0.60
    å·
    -0.60
     Phill
    -0.58
    POSITIVE LOGITS
     necessarily
    1.20
     merely
    1.14
    withstanding
    1.00
     simply
    0.86
     only
    0.82
     mere
    0.82
    icably
    0.81
     just
    0.79
     solely
    0.78
     malice
    0.75
    Activations Density 0.117%

    No Known Activations