© 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 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. 6205
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the presence or absence of conflicts of interest

    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
    314
    +0.17
    0.8%
    256
    +0.12
    0.6%
    729
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    20
    +0.17
    0.09
    455
    +0.12
    0.07
    5
    +0.12
    0.05
    Negative Logits
     Maid
    -0.90
     metaphors
    -0.77
     tropes
    -0.77
     clich
    -0.76
     pedest
    -0.73
     nerds
    -0.72
     buddies
    -0.72
     tricks
    -0.72
     Godd
    -0.72
     nob
    -0.72
    POSITIVE LOGITS
    currently
    0.91
    roximately
    0.88
    imposed
    0.85
     Approximately
    0.83
    iminary
    0.77
    inion
    0.76
    lyak
    0.76
    è£ħ
    0.76
    served
    0.75
    stantial
    0.75
    Activations Density 0.289%

    No Known Activations