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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & 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. 8361
    Prev
    Next
    INDEX
    Explanations

    statements related to government policies and public services

    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
    459
    +0.11
    0.5%
    481
    +0.11
    0.5%
    86
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.11
    0.07
    393
    +0.11
    0.09
    756
    +0.10
    -0.03
    Negative Logits
    colored
    -0.89
    defense
    -0.84
    avored
    -0.82
    ¶
    -0.79
    Los
    -0.76
    ADVERTISEMENT
    -0.74
    Defense
    -0.74
    rollment
    -0.74
    Hide
    -0.73
    favorite
    -0.73
    POSITIVE LOGITS
     realise
    1.53
     recognise
    1.48
     recognised
    1.29
     Ukip
    1.26
     apologise
    1.25
     Nato
    1.17
     emphas
    1.17
     realised
    1.16
     recogn
    1.15
     util
    1.15
    Activations Density 1.075%

    No Known Activations