© 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. 9674
    Prev
    Next
    INDEX
    Explanations

    phrases related to suggestions or implications

    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
    195
    +0.15
    0.7%
    101
    +0.12
    0.5%
    276
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    195
    +0.15
    0.03
    236
    +0.12
    0.04
    274
    +0.10
    0.03
    Negative Logits
    fare
    -0.76
    gie
    -0.76
    kai
    -0.72
    mb
    -0.71
    pes
    -0.71
    tes
    -0.70
    ma
    -0.69
    cod
    -0.68
    train
    -0.68
    Boy
    -0.68
    POSITIVE LOGITS
    ively
    0.94
     otherwise
    0.84
    ibility
    0.74
     hypot
    0.72
    ered
    0.69
     alternatives
    0.68
    ible
    0.67
     explanations
    0.66
     that
    0.66
     reconsider
    0.65
    Activations Density 0.101%

    No Known Activations