© 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. 7411
    Prev
    Next
    INDEX
    Explanations

    phrases related to communication and expressing thoughts or opinions

    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
    332
    +0.14
    0.6%
    380
    +0.11
    0.5%
    487
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    274
    +0.14
    0.03
    332
    +0.11
    0.03
    380
    +0.09
    0.02
    Negative Logits
    Ranked
    -0.81
    ATIONAL
    -0.69
    enegger
    -0.69
    OGR
    -0.68
    MENT
    -0.67
    mental
    -0.67
    OSS
    -0.65
    mercial
    -0.64
    ritional
    -0.64
    Alert
    -0.64
    POSITIVE LOGITS
     goodbye
    1.14
     aloud
    0.87
     hello
    0.79
    dylib
    0.71
     sarcast
    0.70
     farewell
    0.67
    escription
    0.66
     loudly
    0.65
    seless
    0.64
    lihood
    0.64
    Activations Density 0.065%

    No Known Activations