© 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. 537
    Prev
    Next
    INDEX
    Explanations

    words related to rock and metal music genres

    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
    288
    +0.14
    0.6%
    401
    +0.11
    0.5%
    447
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.07
    401
    +0.11
    0.06
    620
    +0.09
    0.02
    Negative Logits
    ciating
    -0.76
    20439
    -0.71
    ividual
    -0.71
     Volunte
    -0.70
    lease
    -0.69
    andum
    -0.68
    ilage
    -0.67
     Rove
    -0.66
    urion
    -0.66
     Estimates
    -0.65
    POSITIVE LOGITS
     genre
    1.27
     genres
    1.19
     thriller
    1.16
     classics
    1.12
     comedy
    1.11
     tropes
    1.11
     opera
    1.05
     revival
    1.04
     punk
    1.03
     realism
    1.01
    Activations Density 0.384%

    No Known Activations