© 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. 1-RES-JB
    11. 5137
    Prev
    Next
    INDEX
    Explanations

    text fragments that contain the character 'ä' at a high activation level

    oai_token-act-pair · gpt-3.5-turbo

    instances of a specific character or symbol used consistently throughout the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ORED
    -0.87
     Jericho
    -0.70
     Gorge
    -0.67
     Mayweather
    -0.65
     Cavaliers
    -0.64
     Sussex
    -0.63
    ored
    -0.61
     rooting
    -0.59
     BlackBerry
    -0.59
     rapt
    -0.59
    POSITIVE LOGITS
    ä
    1.68
    inen
    1.19
    ö
    1.09
    ternity
    1.01
    ka
    0.94
    ü
    0.92
    ¢
    0.91
    onte
    0.89
    î
    0.87
    Ã
    0.87
    Activations Density 0.008%

    No Known Activations