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

    Neuronpedia

    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. 45554
    Prev
    Next
    INDEX
    Explanations

    terms related to severity or intensity, particularly in negative contexts

    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
    453
    +0.13
    0.6%
    411
    +0.12
    0.5%
    730
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    453
    +0.13
    0.01
    411
    +0.12
    0.01
    730
    +0.11
    0.01
    Negative Logits
    ividual
    -0.90
    ovember
    -0.85
    assies
    -0.84
    ĸļ
    -0.81
    akeru
    -0.75
    isphere
    -0.72
    ICLE
    -0.71
    xxxxxxxx
    -0.71
    iliary
    -0.71
    phis
    -0.71
    POSITIVE LOGITS
     punishments
    1.06
     punishment
    1.01
     winters
    1.00
    ness
    0.99
    est
    0.98
    nesses
    0.95
     retribution
    0.95
     harsh
    0.92
     penalties
    0.88
     harshly
    0.87
    Activations Density 0.013%

    No Known Activations