© 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 12288
    9. 8-RES_FS12288-JB
    11. 1253
    Prev
    Next
    INDEX
    Explanations

    informational content like news reports or emergency situations

    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
    347
    +0.11
    0.5%
    196
    +0.11
    0.5%
    333
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    738
    +0.11
    0.02
    347
    +0.11
    0.03
    196
    +0.11
    0.02
    Negative Logits
    ItemTracker
    -0.83
    nir
    -0.79
    ledged
    -0.79
    ©¶æ¥µ
    -0.75
    ento
    -0.71
    afort
    -0.68
    Ħ¢
    -0.68
    Ĥİ
    -0.66
    ²¾
    -0.64
    ucket
    -0.64
    POSITIVE LOGITS
     unfold
    1.44
     videos
    1.08
     replay
    1.00
     VIDEOS
    0.98
     livestream
    0.96
     video
    0.95
     tapes
    0.95
     closely
    0.94
    ideos
    0.92
     documentaries
    0.90
    Activations Density 0.244%

    No Known Activations