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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 4509
    Prev
    Next
    INDEX
    Explanations

    phrases related to political or ideological criticisms

    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
    503
    +0.14
    0.6%
    331
    +0.12
    0.5%
    311
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    331
    +0.14
    0.02
    503
    +0.12
    0.01
    137
    +0.10
    0.01
    Negative Logits
     Madness
    -0.75
     Clause
    -0.72
     Mata
    -0.72
     Tile
    -0.71
     Alma
    -0.71
     Thornton
    -0.68
     Floor
    -0.67
     initials
    -0.67
     Opera
    -0.67
    hyde
    -0.66
    POSITIVE LOGITS
    athom
    1.48
    riend
    1.41
    ortunately
    1.31
    avour
    1.30
    ortun
    1.27
    ortunate
    1.20
    ixed
    1.17
    requently
    1.15
    requ
    1.13
    avored
    1.11
    Activations Density 0.008%

    No Known Activations