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

    Neuronpedia

    Jacobian LensNEW
    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 24576
    9. 8-RES_FS24576-JB
    11. 10352
    Prev
    Next
    INDEX
    Explanations

    terms related to politics, social issues, and justice

    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
    480
    +0.18
    0.8%
    447
    +0.12
    0.6%
    655
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    655
    +0.18
    0.02
    524
    +0.12
    0.02
    168
    +0.11
    0.02
    Negative Logits
    raints
    -0.93
     manif
    -0.89
     horizont
    -0.85
     Instr
    -0.84
     womb
    -0.80
     philos
    -0.78
     misunder
    -0.76
     tentacles
    -0.75
    ngth
    -0.75
     indo
    -0.74
    POSITIVE LOGITS
    âĶĢâĶĢ
    1.06
    âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ
    1.05
    ãĥ¼ãĥ
    1.00
    ishable
    0.99
    à©
    0.93
    cffffcc
    0.90
    à¨
    0.90
    ļ
    0.89
    Torrent
    0.87
    ¤
    0.86
    Activations Density 0.026%

    No Known Activations