© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 172
    Prev
    Next
    INDEX
    Explanations

    phrases related to specific events or incidents, possibly involving law enforcement

    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
    41
    +0.14
    0.6%
    481
    +0.11
    0.5%
    349
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    408
    +0.14
    0.07
    362
    +0.11
    0.06
    476
    +0.10
    0.03
    Negative Logits
    erenn
    -0.84
    sonian
    -0.77
    common
    -0.71
    ancial
    -0.70
    requires
    -0.67
    tle
    -0.65
    friend
    -0.62
    ilus
    -0.61
     improves
    -0.60
    ouble
    -0.59
    POSITIVE LOGITS
     filming
    0.84
     incident
    0.83
     initial
    0.80
     booking
    0.77
     altercation
    0.75
     fateful
    0.74
     recording
    0.73
     deliberations
    0.70
     arra
    0.70
     shooting
    0.70
    Activations Density 0.615%

    No Known Activations