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. 4212
    Prev
    Next
    INDEX
    Explanations

    expressions of obligation, apology, and relationships

    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
    199
    +0.10
    0.5%
    193
    +0.10
    0.5%
    383
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.10
    0.04
    605
    +0.10
    0.04
    193
    +0.10
    0.03
    Negative Logits
    oros
    -0.68
    anova
    -0.67
    ahu
    -0.66
    dra
    -0.66
    tsky
    -0.65
    rake
    -0.65
    puters
    -0.64
    etheless
    -0.64
    arnaev
    -0.64
    ryu
    -0.63
    POSITIVE LOGITS
    sein
    0.71
     manifest
    0.67
     ejac
    0.64
     appl
    0.64
    eming
    0.63
    ¿½
    0.61
     Tob
    0.61
    ©¶æ¥µ
    0.61
     Barth
    0.60
     detriment
    0.60
    Activations Density 0.787%

    No Known Activations