© 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. 1264
    Prev
    Next
    INDEX
    Explanations

    information related to technology and politics

    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
    55
    +0.12
    0.5%
    502
    +0.11
    0.5%
    224
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    502
    +0.12
    0.07
    139
    +0.11
    0.07
    586
    +0.10
    0.07
    Negative Logits
     cubic
    -0.71
    Reviewer
    -0.69
     guiActiveUnfocused
    -0.67
     possession
    -0.64
    violent
    -0.63
    pass
    -0.61
    tones
    -0.60
    Bound
    -0.60
    venge
    -0.59
     survival
    -0.59
    POSITIVE LOGITS
     cites
    1.13
     believes
    1.08
    zbollah
    1.07
     cited
    1.03
     added
    1.03
     cautioned
    1.02
     likened
    1.02
    resy
    1.02
     explained
    1.00
     thinks
    0.99
    Activations Density 0.264%

    No Known Activations