© 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. 11938
    Prev
    Next
    INDEX
    Explanations

    words related to marquee or prestigious elements

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "mar" and its variations in different contexts

    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
    uyomi
    -0.80
    enhagen
    -0.77
    éĹĺ
    -0.72
    itsch
    -0.68
    OLOGY
    -0.67
    ngth
    -0.65
    ãĤĮ
    -0.65
    âĶģ
    -0.64
    mble
    -0.63
    arantine
    -0.62
    POSITIVE LOGITS
    iage
    1.12
    quet
    1.07
    ried
    1.02
    chers
    1.01
    riages
    0.97
    cus
    0.94
    ques
    0.92
    bles
    0.92
    riage
    0.91
    inated
    0.91
    Activations Density 0.020%

    No Known Activations