© 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. 4-RES-JB
    11. 12406
    Prev
    Next
    INDEX
    Explanations

    instances of the word "same" in the text

    oai_token-act-pair · gpt-3.5-turbo

    the repetition of the concept of sameness

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

    No Comments

    Negative Logits
    *=-
    -0.79
     Provided
    -0.65
    omics
    -0.64
    ases
    -0.63
    åĪ
    -0.62
    ç«
    -0.62
    bane
    -0.61
    Å
    -0.61
    ostic
    -0.60
     Laksh
    -0.60
    POSITIVE LOGITS
     thing
    0.97
     exact
    0.85
     amount
    0.84
    rity
    0.77
     vein
    0.75
     kind
    0.73
    nesota
    0.72
     ol
    0.71
    iating
    0.69
     day
    0.69
    Activations Density 0.043%

    No Known Activations