© 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 49152
    9. 8-RES_FS49152-JB
    11. 37243
    Prev
    Next
    INDEX
    Explanations

    dates and times mentioned in the text

    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
    254
    +0.10
    0.4%
    283
    +0.09
    0.4%
    757
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.10
    0.06
    306
    +0.09
    0.05
    114
    +0.09
    0.04
    Negative Logits
    ounge
    -0.75
     hypocr
    -0.68
    cient
    -0.67
    okin
    -0.67
    bred
    -0.65
    ldon
    -0.65
     personality
    -0.64
     comr
    -0.63
     Compan
    -0.62
    ãĤ¤ãĥĪ
    -0.61
    POSITIVE LOGITS
     afternoon
    1.15
     deadline
    0.99
     evening
    0.98
     morning
    0.98
    morning
    0.87
     night
    0.84
     semester
    0.81
     mornings
    0.81
     marking
    0.76
     noon
    0.76
    Activations Density 0.123%

    No Known Activations