© 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 3072
    9. 8-RES_FS3072-JB
    11. 2974
    Prev
    Next
    INDEX
    Explanations

    words related to off-screen events or issues

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    373
    +0.13
    0.6%
    696
    +0.10
    0.5%
    745
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    696
    +0.13
    0.05
    533
    +0.10
    0.04
    568
    +0.10
    0.03
    Negative Logits
    ļéĨĴ
    -0.68
    Untitled
    -0.63
    Export
    -0.61
     Occupations
    -0.60
    ¶ħ
    -0.59
    Ĥİ
    -0.59
    Ĥª
    -0.58
    ãĤ±
    -0.58
    berus
    -0.57
    anners
    -0.57
    POSITIVE LOGITS
    shoot
    1.09
    loaded
    1.09
    site
    1.04
    beat
    1.02
    screen
    0.97
    loading
    0.94
    shore
    0.93
    spr
    0.91
    sites
    0.90
    ices
    0.88
    Activations Density 0.259%

    No Known Activations