© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 1862
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the main message or purpose of a text

    oai_token-act-pair · gpt-3.5-turbo

    the concept of "the point" or main idea in a text

    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
     notor
    -0.83
    uthor
    -0.77
    DAQ
    -0.76
    reditary
    -0.75
    annis
    -0.75
    eco
    -0.73
    phalt
    -0.73
    apons
    -0.72
     tradem
    -0.69
    emale
    -0.68
    POSITIVE LOGITS
    lessly
    0.87
    lessness
    0.86
    point
    0.84
    points
    0.82
     point
    0.72
    posts
    0.71
    blank
    0.71
    ioned
    0.70
     guard
    0.68
    forward
    0.68
    Activations Density 0.032%

    No Known Activations