© 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 24576
    9. 8-RES_FS24576-JB
    11. 11931
    Prev
    Next
    INDEX
    Explanations

    words related to sports awards and achievements in a specific sports conference

    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
    526
    +0.16
    0.8%
    288
    +0.11
    0.5%
    335
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    335
    +0.16
    0.01
    667
    +0.11
    0.01
    187
    +0.10
    0.01
    Negative Logits
     archived
    -0.72
    hower
    -0.70
     beh
    -0.69
     ABE
    -0.68
    âĢ¢âĢ¢
    -0.66
     cataly
    -0.65
    IGH
    -0.64
    éĹĺ
    -0.64
    IGHTS
    -0.63
    UCT
    -0.63
    POSITIVE LOGITS
    emaker
    1.17
    Pac
    1.00
    achu
    0.99
    oco
    0.98
    rice
    0.98
    ific
    0.97
    rome
    0.97
    quet
    0.96
    atown
    0.94
    hes
    0.94
    Activations Density 0.013%

    No Known Activations