© 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 12288
    9. 8-RES_FS12288-JB
    11. 9593
    Prev
    Next
    INDEX
    Explanations

    terms related to corrections or edits in text

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    284
    +0.13
    0.6%
    631
    +0.12
    0.5%
    503
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.13
    0.03
    631
    +0.12
    0.03
    503
    +0.11
    0.02
    Negative Logits
    esses
    -0.71
    ELS
    -0.70
    Premium
    -0.67
    unct
    -0.67
    kefeller
    -0.64
    smith
    -0.63
    ership
    -0.62
    eries
    -0.62
     AMERICA
    -0.61
    emonium
    -0.60
    POSITIVE LOGITS
    asure
    1.30
    rors
    1.16
    aser
    1.07
    asures
    1.06
    bil
    0.99
    rant
    0.97
    lang
    0.95
    asers
    0.95
    got
    0.94
    asing
    0.93
    Activations Density 0.046%

    No Known Activations