© 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. 9349
    Prev
    Next
    INDEX
    Explanations

    numerical values and specific terms related to military operations

    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
    231
    +0.13
    0.6%
    569
    +0.10
    0.5%
    444
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    343
    +0.13
    0.03
    659
    +0.10
    0.03
    538
    +0.10
    0.02
    Negative Logits
    --------------------------------------------------------
    -0.81
     millenn
    -0.68
    ogly
    -0.66
    alty
    -0.66
    ó
    -0.65
    Ú
    -0.65
    Ô
    -0.63
    winner
    -0.63
     contrad
    -0.63
    Else
    -0.63
    POSITIVE LOGITS
    IRT
    1.11
    iolet
    1.11
    III
    1.04
    arsity
    1.04
    ideos
    0.99
    isible
    0.94
    OY
    0.92
    irus
    0.90
    ascular
    0.90
    ancouver
    0.89
    Activations Density 0.040%

    No Known Activations