© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 543
    Prev
    Next
    INDEX
    Explanations

    information related to transportation infrastructure and city development

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    481
    +0.14
    0.6%
    465
    +0.10
    0.4%
    761
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    761
    +0.14
    0.07
    344
    +0.10
    0.08
    608
    +0.10
    0.09
    Negative Logits
    POSE
    -0.71
    PLIED
    -0.69
    Aim
    -0.69
    cus
    -0.68
    reference
    -0.67
    Wing
    -0.65
    TAG
    -0.64
    subject
    -0.64
     Squadron
    -0.64
    ctrl
    -0.62
    POSITIVE LOGITS
     population
    1.07
     inhabitants
    1.01
    population
    0.95
    Population
    0.94
     skyline
    0.89
     populations
    0.89
     coastline
    0.88
     residents
    0.87
     UNESCO
    0.86
     thriving
    0.84
    Activations Density 1.039%

    No Known Activations