© 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 12288
    9. 8-RES_FS12288-JB
    11. 1498
    Prev
    Next
    INDEX
    Explanations

    technology-related terms and phrases

    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
    378
    +0.10
    0.4%
    709
    +0.09
    0.4%
    590
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    115
    +0.10
    0.08
    547
    +0.09
    0.06
    224
    +0.09
    0.08
    Negative Logits
     hath
    -0.54
     proved
    -0.53
     Whilst
    -0.52
    tumblr
    -0.51
     unquestion
    -0.51
     penet
    -0.51
     tho
    -0.50
     terror
    -0.50
     matured
    -0.49
     deval
    -0.49
    POSITIVE LOGITS
    >.
    0.97
    .).
    0.85
    ].
    0.77
    _.
    0.76
    }.
    0.75
    ).
    0.75
    .</
    0.74
    .''
    0.71
    .''.
    0.71
    .]
    0.69
    Activations Density 1.105%

    No Known Activations