© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 20032
    Prev
    Next
    INDEX
    Explanations

    the word "on" repeated

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    373
    +0.12
    0.5%
    718
    +0.10
    0.5%
    543
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    79
    +0.12
    0.02
    422
    +0.10
    0.01
    355
    +0.10
    0.01
    Negative Logits
     Pyr
    -0.74
    itives
    -0.70
    rors
    -0.68
    çĶ°
    -0.66
    ython
    -0.66
     Electoral
    -0.64
    ®
    -0.62
    POS
    -0.61
     Nebula
    -0.61
    ocene
    -0.60
    POSITIVE LOGITS
    shore
    0.90
    uph
    0.89
    rir
    0.83
     ticking
    0.78
     hemor
    0.77
     progressing
    0.77
     pace
    0.76
     emitting
    0.75
     destro
    0.73
     tread
    0.72
    Activations Density 0.036%

    No Known Activations