© 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 24576
    9. 8-RES_FS24576-JB
    11. 9354
    Prev
    Next
    INDEX
    Explanations

    phrases related to politics and government

    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
    259
    +0.11
    0.5%
    431
    +0.10
    0.4%
    689
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    259
    +0.11
    0.03
    431
    +0.10
    0.03
    699
    +0.09
    0.03
    Negative Logits
    ILE
    -0.63
     Mans
    -0.61
    911
    -0.60
    opian
    -0.59
     Petroleum
    -0.59
    anches
    -0.58
     host
    -0.58
    ãĥ©ãĥ³
    -0.57
    hor
    -0.56
     corridors
    -0.55
    POSITIVE LOGITS
    chy
    1.26
    alian
    0.98
    unes
    0.90
     anyways
    0.83
    self
    0.83
    iner
    0.80
     anyway
    0.80
     shipped
    0.78
     happening
    0.77
     happen
    0.77
    Activations Density 0.093%

    No Known Activations