© 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 EvalsExportsinterp-engineNEWAPI 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. 4542
    Prev
    Next
    INDEX
    Explanations

    verbs and adjectives related to imposing regulations or restrictions

    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
    301
    +0.12
    0.5%
    85
    +0.11
    0.5%
    300
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    301
    +0.12
    0.02
    300
    +0.11
    0.02
    85
    +0.10
    0.02
    Negative Logits
    razil
    -0.76
    tes
    -0.74
    wagen
    -0.74
    sey
    -0.73
    vironment
    -0.72
    obby
    -0.72
    abet
    -0.72
    rooms
    -0.72
    say
    -0.71
    rous
    -0.71
    POSITIVE LOGITS
     stricter
    1.06
     sanctions
    1.03
     restrictions
    1.02
     punishments
    0.98
     harsher
    0.97
     burdens
    0.95
     liability
    0.94
     penalties
    0.93
     upon
    0.92
     tariffs
    0.92
    Activations Density 0.041%

    No Known Activations