© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 4758
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and their contextual significance in statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.14
    0.6%
    121
    +0.13
    0.6%
    264
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.14
    0.19
    640
    +0.13
    0.15
    430
    +0.10
    0.14
    Negative Logits
    ttes
    -0.79
    ulhu
    -0.68
     unlocks
    -0.67
     unlock
    -0.67
    chrom
    -0.65
     schedules
    -0.65
    uploads
    -0.64
     taps
    -0.63
     refreshed
    -0.63
    iencies
    -0.61
    POSITIVE LOGITS
     Deborah
    0.94
     Rabbi
    0.87
     Judith
    0.86
     Laura
    0.85
     Kathleen
    0.84
     Carolyn
    0.84
     Roger
    0.83
     Bernard
    0.82
     Heather
    0.81
     Angelo
    0.81
    Activations Density 1.427%

    No Known Activations