© 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. 42974
    Prev
    Next
    INDEX
    Explanations

    instances of the preposition "in."

    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
    640
    +0.15
    0.7%
    123
    +0.13
    0.6%
    11
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.15
    0.06
    469
    +0.13
    0.05
    11
    +0.11
    0.04
    Negative Logits
    henko
    -0.76
    TPP
    -0.76
     vectors
    -0.75
    oeuv
    -0.72
    MpServer
    -0.68
    Animation
    -0.67
    /(
    -0.66
     inval
    -0.66
    acters
    -0.63
     outweigh
    -0.63
    POSITIVE LOGITS
     downtown
    1.19
     suburban
    1.10
     southwest
    1.09
     northwest
    1.06
     Arlington
    1.04
     southwestern
    1.04
     Somerset
    1.03
     southeastern
    1.02
     Melbourne
    1.02
     southeast
    1.01
    Activations Density 0.102%

    No Known Activations