© 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. 26606
    Prev
    Next
    INDEX
    Explanations

    conditional phrases and inquiries about beliefs or opinions

    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
    43
    +0.11
    0.5%
    738
    +0.09
    0.4%
    719
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.11
    0.02
    738
    +0.09
    0.01
    583
    +0.09
    0.01
    Negative Logits
    ãĤ´ãĥ³
    -0.92
    igslist
    -0.91
    âķIJâķIJ
    -0.84
    auri
    -0.81
     guiActiveUn
    -0.81
    Sense
    -0.79
    ãĥĦ
    -0.77
    ionic
    -0.75
    etheus
    -0.75
    Ģ
    -0.73
    POSITIVE LOGITS
     anymore
    0.81
     outcome
    0.72
     anybody
    0.71
    lessly
    0.69
     YOUR
    0.66
     YOU
    0.66
     Rosenstein
    0.65
     Freddie
    0.64
    DERR
    0.64
     ya
    0.63
    Activations Density 0.063%

    No Known Activations