© 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. 22400
    Prev
    Next
    INDEX
    Explanations

    verbs or adjectives related to change or transition

    oai_token-act-pair · gpt-3.5-turbo

    references to transitions or changes in context or state

    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
    mask
    -0.78
    quote
    -0.77
    vic
    -0.75
    iosyn
    -0.74
    VIDEO
    -0.73
    blance
    -0.72
    QUI
    -0.69
    ilver
    -0.66
    Muslim
    -0.66
    insula
    -0.65
    POSITIVE LOGITS
     afar
    1.11
     whence
    0.95
     thence
    0.92
     obscurity
    0.84
     abroad
    0.73
     somewhere
    0.73
     inception
    0.72
     Gutenberg
    0.70
     kindergarten
    0.70
     wherever
    0.67
    Activations Density 0.131%

    No Known Activations