© 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. 8069
    Prev
    Next
    INDEX
    Explanations

    phrases related to unlocking or gaining access to content and achievements in a game

    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
    496
    +0.18
    0.8%
    343
    +0.11
    0.5%
    183
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    496
    +0.18
    0.02
    592
    +0.11
    0.01
    326
    +0.11
    0.02
    Negative Logits
    cats
    -0.70
    ogram
    -0.68
    cas
    -0.68
    ^^^^
    -0.67
    ograms
    -0.66
    mos
    -0.66
     jerk
    -0.64
     jer
    -0.64
    geist
    -0.64
    obar
    -0.61
    POSITIVE LOGITS
     unlocked
    0.79
     unlocking
    0.78
    ledged
    0.75
     Unlock
    0.74
    ibility
    0.73
    oppable
    0.72
    mental
    0.71
    ables
    0.70
    ilts
    0.70
     badges
    0.70
    Activations Density 0.100%

    No Known Activations