© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 8730
    Prev
    Next
    INDEX
    Explanations

    numeric data related to statistics and measurements

    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
    636
    +0.10
    0.5%
    499
    +0.10
    0.5%
    581
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.10
    -0.03
    316
    +0.10
    0.04
    6
    +0.10
    0.05
    Negative Logits
    Reloaded
    -0.75
    ospace
    -0.71
     signalling
    -0.67
    -+-+
    -0.66
     Catalyst
    -0.65
     volley
    -0.65
    "))
    -0.62
     maxim
    -0.61
    Repe
    -0.60
    RAG
    -0.59
    POSITIVE LOGITS
    umbn
    0.78
     Staten
    0.69
     borough
    0.69
    state
    0.66
    Zone
    0.64
     è£ıè
    0.64
    fecture
    0.63
     Streets
    0.63
    insula
    0.61
     ranked
    0.61
    Activations Density 0.291%

    No Known Activations