© 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 49152
    9. 8-RES_FS49152-JB
    11. 37168
    Prev
    Next
    INDEX
    Explanations

    terms related to arrogance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    447
    +0.16
    0.8%
    288
    +0.14
    0.6%
    139
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    381
    +0.16
    0.01
    288
    +0.14
    0.01
    139
    +0.10
    0.01
    Negative Logits
    Downloadha
    -0.82
    cedented
    -0.79
    ajor
    -0.77
    ondo
    -0.77
    EVA
    -0.77
    Interstitial
    -0.76
    OTOS
    -0.75
    heny
    -0.75
    FTWARE
    -0.75
    foreseen
    -0.74
    POSITIVE LOGITS
     arrogance
    1.18
     arrogant
    0.99
     prick
    0.97
     condesc
    0.94
     narciss
    0.90
     smug
    0.89
     asshole
    0.88
     arrog
    0.88
     hypoc
    0.86
     disregard
    0.84
    Activations Density 0.017%

    No Known Activations