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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 18577
    Prev
    Next
    INDEX
    Explanations

    punctuation marks used in the middle of sentences, specifically commas

    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
    393
    +0.10
    0.5%
    497
    +0.09
    0.4%
    430
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    497
    +0.10
    0.03
    289
    +0.09
    0.01
    45
    +0.09
    0.02
    Negative Logits
    RM
    -0.74
    grass
    -0.66
    Contribut
    -0.64
    Executive
    -0.63
    SHIP
    -0.63
    Muslims
    -0.63
    File
    -0.62
    WARE
    -0.62
    Americans
    -0.61
     Writers
    -0.61
    POSITIVE LOGITS
     etc
    1.31
    etc
    1.05
    ...)
    0.86
    â̦)
    0.83
     ...
    0.71
    ĪĴ
    0.71
     â̦
    0.69
     &
    0.68
     respectively
    0.68
     teasp
    0.65
    Activations Density 0.081%

    No Known Activations