© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4846
    Prev
    Next
    INDEX
    Explanations

    instances of the word "wake" and its variations

    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
    414
    +0.12
    0.6%
    343
    +0.11
    0.5%
    526
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    414
    +0.12
    0.02
    372
    +0.11
    0.01
    178
    +0.10
    0.02
    Negative Logits
    regate
    -0.64
    cer
    -0.64
    SHIP
    -0.63
     Proposition
    -0.62
     mileage
    -0.62
    abwe
    -0.61
    entric
    -0.60
    inates
    -0.59
     Forbidden
    -0.59
     Hezbollah
    -0.58
    POSITIVE LOGITS
    mares
    1.19
    fulness
    1.04
    locks
    0.91
    mare
    0.87
    up
    0.85
    own
    0.83
    lock
    0.83
     refreshed
    0.80
    rooms
    0.80
    upt
    0.79
    Activations Density 0.065%

    No Known Activations