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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 46099
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "by" along with its contextual usage

    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
    616
    +0.13
    0.6%
    373
    +0.11
    0.5%
    752
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.13
    0.04
    29
    +0.11
    0.03
    257
    +0.10
    0.03
    Negative Logits
    idad
    -0.65
    heed
    -0.64
    asive
    -0.64
    llor
    -0.63
    ivia
    -0.63
    gin
    -0.57
    istan
    -0.57
     retri
    -0.57
    ocate
    -0.56
    ounter
    -0.56
    POSITIVE LOGITS
     virtue
    1.19
    products
    1.09
     default
    1.08
     implication
    0.97
    product
    0.89
     comparison
    0.88
     leaps
    0.87
    laws
    0.85
    gone
    0.83
    election
    0.81
    Activations Density 0.066%

    No Known Activations