© 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. 5530
    Prev
    Next
    INDEX
    Explanations

    numerical values, particularly IP addresses and related formats or patterns

    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
    +0.15
    0.7%
    586
    +0.10
    0.5%
    72
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.15
    0.02
    107
    +0.10
    0.02
    63
    +0.10
    0.02
    Negative Logits
    âĹ¼
    -0.75
     batches
    -0.65
     unmatched
    -0.65
     ingested
    -0.65
     brakes
    -0.64
    itted
    -0.63
     MSI
    -0.60
     ingestion
    -0.58
     mornings
    -0.58
     aval
    -0.57
    POSITIVE LOGITS
    aden
    0.74
    gross
    0.73
    vantage
    0.71
    abba
    0.71
    amus
    0.69
    bia
    0.68
    abit
    0.65
    ymes
    0.65
    xp
    0.65
    ilar
    0.64
    Activations Density 0.156%

    No Known Activations