Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 995
    Prev
    Next
    INDEX
    Explanations

    player-related terms or mentions

    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
    575
    +0.14
    0.6%
    447
    +0.14
    0.6%
    216
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.14
    0.05
    189
    +0.14
    0.04
    241
    +0.12
    0.04
    Negative Logits
    ories
    -0.88
    Operation
    -0.88
    ĺħ
    -0.71
    ħĭ
    -0.70
    ½
    -0.69
     Cooking
    -0.68
    ¿½
    -0.66
    een
    -0.65
    orship
    -0.64
    cess
    -0.64
    POSITIVE LOGITS
     drafted
    0.84
     endors
    0.83
     kneeling
    0.80
     jerseys
    0.79
    unknown
    0.72
     rosters
    0.71
    player
    0.71
     athlete
    0.70
     Mahmoud
    0.70
    inic
    0.69
    Activations Density 0.117%

    No Known Activations