Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 41383
    Prev
    Next
    INDEX
    Explanations

    nouns and terms related to technology, entertainment, and organization

    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
    447
    +0.19
    0.9%
    288
    +0.17
    0.8%
    481
    +0.15
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.19
    0.08
    575
    +0.17
    0.06
    40
    +0.15
    0.04
    Negative Logits
    inces
    -0.74
     Occupations
    -0.74
    tics
    -0.73
     Empires
    -0.71
    places
    -0.68
    idents
    -0.68
    uckles
    -0.67
    sites
    -0.67
     Ones
    -0.67
     Indra
    -0.66
    POSITIVE LOGITS
    less
    0.88
     resembling
    0.79
     consisting
    0.75
    like
    0.75
    ogram
    0.75
     cutter
    0.73
    urable
    0.71
    ful
    0.71
     analogy
    0.69
     sized
    0.69
    Activations Density 0.321%

    No Known Activations