© 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 12288
    9. 8-RES_FS12288-JB
    11. 9873
    Prev
    Next
    INDEX
    Explanations

    phrases related to making or creating something

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    402
    +0.12
    0.6%
    100
    +0.11
    0.5%
    245
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    402
    +0.12
    0.08
    100
    +0.11
    0.07
    245
    +0.11
    0.07
    Negative Logits
    thia
    -0.63
     Niet
    -0.60
    ;;;;
    -0.59
    assis
    -0.59
    hood
    -0.58
    istine
    -0.57
     alluded
    -0.57
    bour
    -0.57
     presided
    -0.56
     Mub
    -0.56
    POSITIVE LOGITS
     sure
    1.37
    hift
    1.09
     adjustments
    0.86
     sense
    0.85
     excuses
    0.76
    solete
    0.76
     AMERICA
    0.76
     Sure
    0.75
     decisions
    0.75
    URE
    0.74
    Activations Density 0.146%

    No Known Activations