© 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. 1671
    Prev
    Next
    INDEX
    Explanations

    sequences of special characters and different character sets like ASCII and Unicode characters

    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
    637
    +0.12
    0.5%
    202
    +0.10
    0.5%
    503
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.12
    0.02
    637
    +0.10
    0.01
    28
    +0.09
    0.01
    Negative Logits
    merce
    -0.67
    xus
    -0.66
     Ples
    -0.63
     Mandela
    -0.61
    gow
    -0.59
    bourg
    -0.59
     intersections
    -0.57
     Woodward
    -0.57
    horn
    -0.56
     closures
    -0.56
    POSITIVE LOGITS
    ľ
    1.10
    Ķ
    1.03
    ´
    1.02
    ł
    1.02
    Ħ
    1.01
    ģ
    0.97
    ¤
    0.96
    ĩ
    0.96
    ĺ
    0.94
    IJ
    0.94
    Activations Density 0.019%

    No Known Activations