© 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. 25749
    Prev
    Next
    INDEX
    Explanations

    recurring patterns or repetitions in terms, particularly focusing on the term "bro"

    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
    481
    +0.14
    0.6%
    250
    +0.11
    0.5%
    533
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    667
    +0.14
    0.02
    232
    +0.11
    0.03
    37
    +0.10
    0.02
    Negative Logits
    ogene
    -0.80
    hered
    -0.80
     disapp
    -0.79
     Julie
    -0.75
     Bauer
    -0.72
    ived
    -0.67
     adul
    -0.65
    IFE
    -0.65
    Ñı
    -0.64
     enlarg
    -0.64
    POSITIVE LOGITS
    t
    1.35
    tz
    1.19
    T
    1.18
    ts
    1.14
    Ts
    1.13
    tty
    1.13
    TD
    1.09
    ti
    1.07
    tis
    1.07
    TS
    1.06
    Activations Density 0.109%

    No Known Activations