© 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 12288
    9. 8-RES_FS12288-JB
    11. 1289
    Prev
    Next
    INDEX
    Explanations

    words related to legal proceedings and criminal activities

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    659
    +0.14
    0.6%
    248
    +0.11
    0.5%
    720
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.08
    167
    +0.11
    0.02
    67
    +0.10
    0.05
    Negative Logits
     pony
    -0.71
     Cerberus
    -0.70
     Costco
    -0.65
     ALS
    -0.64
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.63
     CBI
    -0.63
     Nirvana
    -0.62
     Sno
    -0.62
    EStream
    -0.62
    pherd
    -0.61
    POSITIVE LOGITS
    ifully
    1.23
    itting
    1.22
    rarily
    1.20
    ately
    1.16
    aring
    1.15
    cially
    1.14
    ounding
    1.12
    ipping
    1.11
    cknowled
    1.10
    lying
    1.10
    Activations Density 0.363%

    No Known Activations