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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 5548
    Prev
    Next
    INDEX
    Explanations

    terms indicating specific locations or contexts

    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
    739
    +0.10
    0.4%
    230
    +0.10
    0.4%
    171
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    299
    +0.10
    0.06
    575
    +0.10
    0.06
    27
    +0.09
    0.04
    Negative Logits
    iffs
    -0.82
    lations
    -0.70
    phia
    -0.69
    hetti
    -0.66
    aido
    -0.65
    alities
    -0.64
    yrights
    -0.64
    gets
    -0.63
    etti
    -0.63
    oulos
    -0.62
    POSITIVE LOGITS
    Ĵ
    0.82
     unrelated
    0.78
     unspecified
    0.77
    SPONSORED
    0.76
     inaccessible
    0.74
     hostile
    0.71
     supposedly
    0.71
     inhabited
    0.70
     governed
    0.69
     deemed
    0.68
    Activations Density 0.555%

    No Known Activations