© 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. 6487
    Prev
    Next
    INDEX
    Explanations

    special characters in non-English alphabets that imply a specific character or meaning

    oai_token-act-pair · gpt-3.5-turbo

    appearances of the character "ŀ"

    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
    pton
    -0.75
     disenfranch
    -0.70
     hemor
    -0.69
    abwe
    -0.69
    ufact
    -0.68
    Downloadha
    -0.68
     metic
    -0.67
     reperc
    -0.67
     vulner
    -0.65
     philos
    -0.63
    POSITIVE LOGITS
    ŀ
    1.10
    Ĺ
    1.01
    IJ
    0.91
    ĵ
    0.89
    ĭ
    0.89
    ł
    0.89
    ³
    0.89
    ·
    0.89
    ļ
    0.88
    ĺ
    0.86
    Activations Density 0.006%

    No Known Activations