© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 4122
    Prev
    Next
    INDEX
    Explanations

    various forms of the word "est" or words related to status or classification

    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
    21
    +0.11
    0.5%
    551
    +0.11
    0.5%
    335
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.11
    0.02
    701
    +0.11
    0.01
    19
    +0.10
    0.02
    Negative Logits
    vier
    -0.96
    oot
    -0.90
    usters
    -0.87
    isers
    -0.85
    ounce
    -0.84
    igel
    -0.84
    iture
    -0.84
    icky
    -0.83
    iser
    -0.79
    ichen
    -0.79
    POSITIVE LOGITS
    hetics
    0.76
     Norn
    0.68
    hetic
    0.68
    URA
    0.67
     Freak
    0.67
    rad
    0.64
     Oro
    0.63
     Rosenstein
    0.61
    çĭ
    0.61
     Radiant
    0.61
    Activations Density 0.101%

    No Known Activations