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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 45885
    Prev
    Next
    INDEX
    Explanations

    phrases related to surveillance and security measures against terrorism

    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
    481
    +0.14
    0.6%
    218
    +0.14
    0.6%
    265
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    298
    +0.14
    0.04
    265
    +0.14
    0.04
    686
    +0.10
    0.03
    Negative Logits
    iture
    -0.94
    ĸļ
    -0.87
    cknow
    -0.82
    ruary
    -0.81
    itures
    -0.81
    eton
    -0.80
    endars
    -0.79
    eatures
    -0.75
     Compensation
    -0.73
    åħī
    -0.71
    POSITIVE LOGITS
     malware
    1.40
     pathogens
    1.39
     terrorists
    1.34
     virus
    1.32
     toxins
    1.31
     viruses
    1.29
     infection
    1.28
     terrorist
    1.26
     tuberculosis
    1.21
     toxin
    1.21
    Activations Density 0.556%

    No Known Activations