Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 6-RES_POST_32K-OAI
    6. 4191
    Prev
    Next
    INDEX
    Explanations

    negations related to capability or consent

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 6-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.6.hook_resid_post
    Hook Layer
    6
    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
    70
    +0.11
    0.5%
    643
    +0.10
    0.5%
    122
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    643
    +0.11
    0.05
    551
    +0.10
    0.04
    307
    +0.10
    0.04
    Negative Logits
    estate
    -0.71
    imity
    -0.69
    CI
    -0.68
    rant
    -0.65
    mails
    -0.63
    chwitz
    -0.61
     Appears
    -0.61
    hyde
    -0.60
     Shot
    -0.58
     sidx
    -0.58
    POSITIVE LOGITS
     afford
    1.26
     imagine
    0.88
     feas
    0.87
     withstand
    0.87
     conceive
    0.86
     cope
    0.86
     decipher
    0.83
     muster
    0.81
    berra
    0.80
     possibly
    0.80
    Activations Density 0.072%

    No Known Activations