© 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. 8407
    Prev
    Next
    INDEX
    Explanations

    statements of identity or association

    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
    +0.13
    0.6%
    551
    +0.10
    0.4%
    700
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    668
    +0.13
    0.05
    55
    +0.10
    0.06
    499
    +0.10
    0.04
    Negative Logits
     overwhelming
    -0.63
     dstg
    -0.63
    Party
    -0.63
    addons
    -0.60
    ARP
    -0.59
     overl
    -0.58
    Therefore
    -0.58
     Neh
    -0.57
    LOD
    -0.57
    Whereas
    -0.57
    POSITIVE LOGITS
     QC
    0.91
     Jr
    0.85
     PhD
    0.77
    elaide
    0.74
     (@
    0.72
     Sr
    0.72
    ibaba
    0.67
     alias
    0.65
     captcha
    0.64
    ichen
    0.63
    Activations Density 0.617%

    No Known Activations