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. 41192
    Prev
    Next
    INDEX
    Explanations

    references to the LGBTQ community and related issues

    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
    314
    +0.11
    0.5%
    153
    +0.11
    0.5%
    496
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    147
    +0.11
    0.02
    57
    +0.11
    0.01
    153
    +0.11
    0.01
    Negative Logits
     reper
    -0.76
    frames
    -0.71
     Canaver
    -0.69
     Hitchcock
    -0.68
    Downloadha
    -0.68
     guiActiveUnfocused
    -0.65
    acca
    -0.65
     Rove
    -0.65
    amina
    -0.64
     Manufacturer
    -0.63
    POSITIVE LOGITS
    IQ
    1.01
    IA
    0.87
    UE
    0.85
    WER
    0.82
    LGBT
    0.81
     spectrum
    0.79
    ues
    0.79
     Spectrum
    0.76
     activists
    0.76
    obic
    0.76
    Activations Density 0.009%

    No Known Activations