© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 45601
    Prev
    Next
    INDEX
    Explanations

    references to enclosed spaces or barriers

    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
    59
    +0.14
    0.7%
    288
    +0.13
    0.6%
    205
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.01
    40
    +0.13
    0.01
    516
    +0.12
    0.01
    Negative Logits
    enthal
    -0.84
    ãĥ£
    -0.74
    nant
    -0.74
    76561
    -0.71
    ãĤ§
    -0.70
    rouse
    -0.68
    zinski
    -0.67
    ku
    -0.66
    ruary
    -0.65
    enhagen
    -0.65
    POSITIVE LOGITS
    osures
    1.33
     encl
    0.98
    izabeth
    0.87
     enclosed
    0.83
     enclosure
    0.83
     spaces
    0.76
    ridges
    0.73
     circuits
    0.72
    ously
    0.72
    aves
    0.71
    Activations Density 0.006%

    No Known Activations