© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 43034
    Prev
    Next
    INDEX
    Explanations

    references to political figures and their statements

    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
    481
    +0.14
    0.6%
    725
    +0.13
    0.6%
    762
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.14
    0.06
    139
    +0.13
    0.05
    224
    +0.12
    0.05
    Negative Logits
    geries
    -0.98
    ective
    -0.82
    finder
    -0.77
    iencies
    -0.77
    earch
    -0.74
    ocyte
    -0.72
     ILCS
    -0.72
    abase
    -0.71
    production
    -0.69
     miscarriage
    -0.68
    POSITIVE LOGITS
     remarks
    1.19
     reiterated
    1.17
     praised
    1.16
     spoke
    1.15
     thanked
    1.07
     reiter
    1.06
     said
    1.03
     uttered
    1.03
     reiterate
    1.03
     stressed
    1.01
    Activations Density 0.208%

    No Known Activations