© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 1887
    Prev
    Next
    INDEX
    Explanations

    specific strings of characters or words related to well-known fictional characters, especially from fantasy or science fiction literature 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_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
    152
    +0.12
    0.6%
    523
    +0.10
    0.4%
    635
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.12
    0.02
    523
    +0.10
    0.02
    128
    +0.09
    0.01
    Negative Logits
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.80
    ciation
    -0.70
    eph
    -0.65
    yright
    -0.64
    hyde
    -0.61
     Emirates
    -0.61
     timely
    -0.60
     diminishing
    -0.59
    ei
    -0.59
    peat
    -0.58
    POSITIVE LOGITS
    atton
    1.29
    ategy
    1.25
    anded
    1.16
    ategic
    1.16
    icken
    1.12
    ateg
    1.05
    ife
    1.05
    ands
    1.05
    angers
    1.04
    ings
    1.04
    Activations Density 0.022%

    No Known Activations