© 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. 10-RES-JB
    11. 2915
    Prev
    Next
    INDEX
    Explanations

    direct speech indicating someone speaking to the person referenced as "told me"

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "me" indicating direct communication or personal statements

    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
    Released
    -0.72
    ģĸ
    -0.71
    wake
    -0.64
    inventoryQuantity
    -0.63
    isoft
    -0.59
    ãĤ«
    -0.55
    taboola
    -0.55
    emort
    -0.54
    imize
    -0.54
    Fit
    -0.54
    POSITIVE LOGITS
     privately
    0.91
     bluntly
    0.89
     anecd
    0.86
     beforehand
    0.81
     via
    0.77
     about
    0.75
     plainly
    0.73
     orally
    0.70
    oho
    0.69
     confidently
    0.67
    Activations Density 0.049%

    No Known Activations