© 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. 7328
    Prev
    Next
    INDEX
    Explanations

    phrases and nouns related to geographic locations

    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
    393
    +0.11
    0.5%
    290
    +0.10
    0.5%
    175
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    53
    +0.11
    0.04
    476
    +0.10
    0.03
    611
    +0.10
    0.03
    Negative Logits
    maid
    -0.78
    atu
    -0.73
    ufact
    -0.72
    gyn
    -0.69
     sqor
    -0.68
    ÙĨ
    -0.68
    byn
    -0.66
     destro
    -0.66
    abilities
    -0.66
    oult
    -0.64
    POSITIVE LOGITS
     coastline
    1.14
     territory
    0.93
     roadway
    0.86
     pavement
    0.84
     woodland
    0.83
     grass
    0.80
     airspace
    0.78
     swath
    0.77
     countryside
    0.77
     asphalt
    0.76
    Activations Density 0.111%

    No Known Activations