Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 46143
    Prev
    Next
    INDEX
    Explanations

    personal experiences related to significant life events and relationships

    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
    480
    +0.45
    2.3%
    266
    +0.11
    0.6%
    378
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.45
    0.01
    266
    +0.11
    0.07
    351
    +0.10
    0.03
    Negative Logits
    Alternatively
    -0.63
    imble
    -0.63
    Elsewhere
    -0.62
    xit
    -0.62
    İĭ
    -0.59
    ãĥī
    -0.59
    FIG
    -0.59
    ©¶æ¥µ
    -0.58
    ħĭ
    -0.57
    bris
    -0.56
    POSITIVE LOGITS
     my
    1.08
     myself
    0.82
     haha
    0.80
     fuckin
    0.79
     me
    0.79
    my
    0.76
     kinda
    0.75
     alot
    0.74
     MY
    0.74
     horrible
    0.73
    Activations Density 0.699%

    No Known Activations