© 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 49152
    9. 8-RES_FS49152-JB
    11. 37323
    Prev
    Next
    INDEX
    Explanations

    instances of the word "hear" and its variations, indicating a focus on auditory experiences or related topics

    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
    380
    +0.13
    0.6%
    530
    +0.12
    0.6%
    306
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    380
    +0.13
    0.01
    530
    +0.12
    0.01
    435
    +0.11
    0.01
    Negative Logits
     mortg
    -0.65
     jumper
    -0.61
    igsaw
    -0.60
     paved
    -0.59
     leap
    -0.58
     squid
    -0.58
     wound
    -0.57
     leaps
    -0.56
     sauces
    -0.56
     shroud
    -0.55
    POSITIVE LOGITS
    ings
    1.03
    ths
    0.99
    INGS
    0.94
     Hear
    0.90
     voices
    0.86
    ances
    0.86
     hear
    0.83
    ience
    0.81
    GROUND
    0.77
    rums
    0.76
    Activations Density 0.012%

    No Known Activations