© 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. 841
    Prev
    Next
    INDEX
    Explanations

    words related to crime and investigations

    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
    393
    +0.19
    0.9%
    362
    +0.12
    0.5%
    379
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.19
    0.09
    393
    +0.12
    0.07
    608
    +0.09
    0.07
    Negative Logits
    eers
    -0.81
    arians
    -0.70
    uba
    -0.70
    each
    -0.69
    atoon
    -0.68
    aris
    -0.68
    SPONSORED
    -0.68
     Akin
    -0.65
    ãĥ»
    -0.64
    kell
    -0.64
    POSITIVE LOGITS
     equation
    1.14
     puzzle
    1.05
     larger
    0.91
     same
    0.88
     broader
    0.87
     ecosystem
    0.87
     wider
    0.86
     continuum
    0.85
     workforce
    0.81
     solution
    0.80
    Activations Density 0.332%

    No Known Activations