© 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. 349
    Prev
    Next
    INDEX
    Explanations

    words related to accusations or claims of wrongdoing and investigations

    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
    288
    +0.11
    0.5%
    195
    +0.11
    0.5%
    447
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    395
    +0.11
    0.03
    263
    +0.11
    0.03
    274
    +0.10
    0.03
    Negative Logits
    keys
    -0.90
    arton
    -0.77
     busiest
    -0.75
    oppy
    -0.71
    å¸
    -0.70
    OTS
    -0.68
    skill
    -0.68
    perm
    -0.68
    tz
    -0.67
    peror
    -0.67
    POSITIVE LOGITS
     alleging
    1.15
     leveled
    1.13
     accusing
    1.12
     allegations
    1.10
     against
    1.08
     leve
    1.05
     levied
    1.04
     accusation
    1.03
     accusations
    1.03
     alleges
    0.99
    Activations Density 0.090%

    No Known Activations