© 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. 41612
    Prev
    Next
    INDEX
    Explanations

    occurrences of the suffix 'er' in various 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
    216
    +0.12
    0.6%
    683
    +0.10
    0.4%
    706
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    683
    +0.12
    0.02
    284
    +0.10
    0.02
    373
    +0.09
    -0.01
    Negative Logits
    hift
    -0.78
    ĸļ
    -0.71
    scl
    -0.69
     tremend
    -0.61
    sers
    -0.59
    iesel
    -0.58
    aspers
    -0.58
    s
    -0.57
    gpu
    -0.57
    osate
    -0.56
    POSITIVE LOGITS
    cial
    0.90
    cially
    0.90
    rett
    0.82
    berus
    0.81
    aneous
    0.81
    rors
    0.78
    nel
    0.78
    amic
    0.77
    cul
    0.77
    jee
    0.73
    Activations Density 0.024%

    No Known Activations