© 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. 17755
    Prev
    Next
    INDEX
    Explanations

    information related to nuclear power plants

    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
    143
    +0.11
    0.5%
    343
    +0.10
    0.4%
    18
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.06
    115
    +0.10
    0.04
    41
    +0.10
    0.04
    Negative Logits
     Helpful
    -0.81
     Kinnikuman
    -0.80
    ricular
    -0.70
    urate
    -0.70
    EStream
    -0.69
    Ô
    -0.68
    Pg
    -0.67
     Parenthood
    -0.66
    idious
    -0.65
    IRED
    -0.64
    POSITIVE LOGITS
     facility
    0.86
     scheme
    0.85
     series
    0.82
     program
    0.81
     fiasco
    0.81
     compound
    0.81
     debacle
    0.80
     division
    0.80
     consortium
    0.77
     enclave
    0.77
    Activations Density 0.271%

    No Known Activations