© 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 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. 17780
    Prev
    Next
    INDEX
    Explanations

    references to legal cases and criminal activities involving individuals

    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
    612
    +0.14
    0.6%
    454
    +0.11
    0.5%
    224
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.14
    0.08
    612
    +0.11
    0.08
    224
    +0.11
    0.07
    Negative Logits
    kefeller
    -0.81
    emetery
    -0.72
    antage
    -0.72
    undo
    -0.69
    vernment
    -0.69
     Observatory
    -0.65
    ypes
    -0.64
     Skydragon
    -0.64
     Technical
    -0.64
    perties
    -0.63
    POSITIVE LOGITS
     herself
    1.55
    pher
    1.44
    athed
    1.27
    athing
    1.23
    pard
    1.19
    ffield
    1.09
    pherd
    1.09
    ikh
    1.02
    lled
    1.00
     husband
    0.99
    Activations Density 0.115%

    No Known Activations