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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 20015
    Prev
    Next
    INDEX
    Explanations

    structured information such as identification numbers and plate numbers

    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
    586
    +0.12
    0.5%
    197
    +0.10
    0.5%
    239
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.03
    314
    +0.10
    0.00
    289
    +0.09
    0.01
    Negative Logits
     furious
    -0.79
     mindless
    -0.72
     attrition
    -0.71
     coughing
    -0.71
     closures
    -0.69
     snowball
    -0.68
     healthier
    -0.68
     gloom
    -0.68
    ependence
    -0.68
     temptation
    -0.67
    POSITIVE LOGITS
    20439
    1.29
    003
    1.29
    CV
    1.28
    201
    1.27
    014
    1.21
    XXXX
    1.19
    005
    1.19
    0002
    1.19
    xxxxxxxx
    1.18
    017
    1.15
    Activations Density 0.156%

    No Known Activations