© 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 49152
    9. 8-RES_FS49152-JB
    11. 37373
    Prev
    Next
    INDEX
    Explanations

    variations of the suffix "-ism," indicating ideologies, practices, or belief systems

    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
    328
    +0.12
    0.5%
    497
    +0.11
    0.5%
    701
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    701
    +0.12
    0.01
    464
    +0.11
    0.01
    277
    +0.10
    0.01
    Negative Logits
    vag
    -0.65
    SPONSORED
    -0.65
    Ùĩ
    -0.65
    FACE
    -0.62
    glers
    -0.62
     accord
    -0.60
     bombard
    -0.60
     grown
    -0.60
    VICE
    -0.59
     menstru
    -0.59
    POSITIVE LOGITS
    atic
    1.38
    atics
    1.33
    edia
    1.14
    yip
    1.08
    ophe
    1.05
    atical
    1.04
    achine
    1.02
    atically
    1.01
    useum
    1.00
    agic
    0.97
    Activations Density 0.014%

    No Known Activations