© 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 49152
    9. 8-RES_FS49152-JB
    11. 37428
    Prev
    Next
    INDEX
    Explanations

    repeated mentions or addresses to the reader or audience

    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
    267
    +0.14
    0.6%
    156
    +0.10
    0.4%
    718
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    267
    +0.14
    0.05
    156
    +0.10
    0.04
    49
    +0.09
    0.04
    Negative Logits
    ãĥ³ãĤ¸
    -0.72
    İĭ
    -0.59
     Ambro
    -0.59
     Crusade
    -0.57
    Balt
    -0.57
    MpServer
    -0.57
    Ô
    -0.56
    Bind
    -0.56
     Different
    -0.56
    HCR
    -0.56
    POSITIVE LOGITS
    're
    1.36
    've
    1.17
     wanna
    1.10
     guessed
    1.08
     want
    1.00
     guys
    0.95
     weren
    0.92
    'd
    0.90
     haven
    0.89
     think
    0.89
    Activations Density 0.052%

    No Known Activations