© 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. 9856
    Prev
    Next
    INDEX
    Explanations

    descriptions of whether someone or something should be given deserved outcomes or treatment

    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
    533
    +0.11
    0.5%
    7
    +0.11
    0.5%
    52
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    533
    +0.11
    0.03
    52
    +0.11
    0.02
    13
    +0.10
    0.02
    Negative Logits
    ullivan
    -0.71
    shr
    -0.60
    edd
    -0.60
     cycl
    -0.60
     WI
    -0.60
     plateau
    -0.59
    ming
    -0.58
    ula
    -0.57
    bands
    -0.57
    gdala
    -0.56
    POSITIVE LOGITS
    arna
    0.99
     applause
    0.93
     praise
    0.87
     precedence
    0.85
     attention
    0.84
     consideration
    0.84
     criticism
    0.82
     credit
    0.82
     scrutiny
    0.80
     accol
    0.77
    Activations Density 0.065%

    No Known Activations