© 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 49152
    9. 8-RES_FS49152-JB
    11. 42886
    Prev
    Next
    INDEX
    Explanations

    accent marks or diacritical marks in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    48
    +0.13
    0.6%
    62
    +0.11
    0.5%
    103
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    103
    +0.13
    0.01
    48
    +0.11
    0.00
    107
    +0.11
    0.00
    Negative Logits
    anwhile
    -0.74
     bonded
    -0.73
     positively
    -0.70
     dope
    -0.69
    OHN
    -0.68
     counters
    -0.66
     makers
    -0.66
     cocaine
    -0.65
    staking
    -0.65
     jewelry
    -0.64
    POSITIVE LOGITS
    ĵ
    1.46
    Ĵ
    1.43
    ł
    1.40
    Ł
    1.39
    ĸļ士
    1.35
    ±
    1.34
    ŀ
    1.34
    Ŀ
    1.32
    ĨĴ
    1.32
    ķ
    1.29
    Activations Density 0.001%

    No Known Activations