© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 9106
    Prev
    Next
    INDEX
    Explanations

    sentences related to catastrophic events or global phenomena

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    679
    +0.21
    1.0%
    447
    +0.19
    0.9%
    174
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.21
    0.12
    6
    +0.19
    0.11
    440
    +0.10
    0.08
    Negative Logits
    onite
    -0.77
     mate
    -0.77
     calf
    -0.73
    ebus
    -0.70
    iris
    -0.67
     mates
    -0.66
    uana
    -0.66
    ube
    -0.66
     unspecified
    -0.66
     remorse
    -0.65
    POSITIVE LOGITS
     Whether
    1.29
     Often
    1.21
     Especially
    1.20
     Particularly
    1.17
     Typically
    1.15
     However
    1.14
     Many
    1.11
     Few
    1.10
     Yet
    1.07
     Although
    1.07
    Activations Density 0.437%

    No Known Activations