© 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. 16891
    Prev
    Next
    INDEX
    Explanations

    special characters and codes, potentially related to language translation or data formatting

    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
    +0.12
    0.6%
    276
    +0.12
    0.6%
    699
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    50
    +0.12
    0.01
    503
    +0.12
    0.01
    284
    +0.11
    0.01
    Negative Logits
     Danger
    -0.66
    auga
    -0.62
     anat
    -0.61
    worldly
    -0.59
     lapse
    -0.58
     pige
    -0.58
     Dragonbound
    -0.58
     Loans
    -0.57
    BuyableInstoreAndOnline
    -0.57
    raints
    -0.57
    POSITIVE LOGITS
    Ķ
    1.12
    ł
    1.01
    ¹
    1.00
    âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ
    0.98
    ·
    0.97
    ħ
    0.96
    ¼
    0.96
    а
    0.93
    éĹĺ
    0.92
    IJ
    0.92
    Activations Density 0.012%

    No Known Activations