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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5333
    Prev
    Next
    INDEX
    Explanations

    references to locations and living situations

    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
    613
    +0.13
    0.6%
    687
    +0.11
    0.5%
    431
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    685
    +0.13
    0.04
    665
    +0.11
    0.05
    431
    +0.11
    0.04
    Negative Logits
    Sund
    -0.65
     Lima
    -0.57
     slack
    -0.57
    rane
    -0.57
     Blank
    -0.57
     explan
    -0.57
     Gunn
    -0.57
     Kle
    -0.56
     millenn
    -0.56
     scarce
    -0.55
    POSITIVE LOGITS
    quished
    0.72
    cript
    0.64
    chool
    0.64
    abouts
    0.61
    dn
    0.60
    */(
    0.60
    ignt
    0.60
    ,[
    0.59
    uron
    0.59
    stead
    0.59
    Activations Density 0.648%

    No Known Activations