© 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. 8-RES-JB
    11. 24500
    Prev
    Next
    INDEX
    Explanations

    instances where there is a contradiction or contrast in the text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "despite" and similar phrases indicating contrast or contradiction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    pecially
    -0.84
    orthy
    -0.77
    uran
    -0.77
    ¬¼
    -0.76
    regon
    -0.76
    ãĤ©
    -0.75
    abama
    -0.75
    rew
    -0.74
    urated
    -0.74
    rette
    -0.74
    POSITIVE LOGITS
     nevertheless
    1.24
     still
    1.06
     nonetheless
    1.03
     remains
    0.88
    still
    0.86
     however
    0.82
     alas
    0.82
     neither
    0.80
     there
    0.75
     remain
    0.75
    Activations Density 0.237%

    No Known Activations