© 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 1536
    9. 8-RES_FS1536-JB
    11. 14
    Prev
    Next
    INDEX
    Explanations

    phrases related to superhero characters, comics, and movies

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    447
    +0.12
    0.5%
    738
    +0.10
    0.5%
    701
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.12
    0.15
    326
    +0.10
    0.11
    354
    +0.10
    0.08
    Negative Logits
    ¿½
    -0.80
    ateur
    -0.78
    ntil
    -0.76
    mination
    -0.76
    vant
    -0.72
    gain
    -0.69
    lah
    -0.68
    ickr
    -0.67
    rontal
    -0.67
    rikes
    -0.65
    POSITIVE LOGITS
     Characters
    1.12
     Comics
    1.02
     villains
    1.01
     Reborn
    0.96
     Universe
    0.94
     Adventures
    0.92
    acters
    0.92
     villain
    0.91
     superheroes
    0.90
     Origins
    0.86
    Activations Density 2.302%

    No Known Activations