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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 583
    Prev
    Next
    INDEX
    Explanations

    statements related to political activities and legal proceedings

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    227
    +0.11
    0.5%
    610
    +0.10
    0.5%
    448
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    610
    +0.11
    0.08
    659
    +0.10
    0.10
    224
    +0.10
    0.10
    Negative Logits
    enhagen
    -0.77
     SDL
    -0.68
    solete
    -0.68
    Players
    -0.67
    undo
    -0.66
    PDATE
    -0.66
     shootout
    -0.65
    olkien
    -0.65
     sidx
    -0.65
    ighth
    -0.65
    POSITIVE LOGITS
     herself
    2.16
     husband
    1.20
     Louise
    1.18
     her
    1.15
     hijab
    1.13
     Devi
    1.11
     pregnant
    1.10
     she
    1.08
     breasts
    1.08
     bikini
    1.08
    Activations Density 2.993%

    No Known Activations