© 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 24576
    9. 8-RES_FS24576-JB
    11. 1740
    Prev
    Next
    INDEX
    Explanations

    mentions of pointing out or highlighting specific information

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    236
    +0.12
    0.5%
    642
    +0.11
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    236
    +0.12
    0.02
    642
    +0.11
    0.02
    69
    +0.10
    0.02
    Negative Logits
    wake
    -0.98
     experien
    -0.78
    ogue
    -0.74
    soever
    -0.73
    hei
    -0.73
    isol
    -0.69
    shaw
    -0.68
    respective
    -0.66
    isode
    -0.65
    catentry
    -0.65
    POSITIVE LOGITS
     inconsistencies
    1.01
     how
    0.94
     loopholes
    0.92
     contradictions
    0.90
     flaws
    0.88
     discrepancies
    0.88
     similarities
    0.81
     weaknesses
    0.80
     differences
    0.78
     why
    0.76
    Activations Density 0.027%

    No Known Activations