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

    Neuronpedia

    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. 5771
    Prev
    Next
    INDEX
    Explanations

    occurrences of words related to extremism

    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
    373
    +0.14
    0.6%
    497
    +0.11
    0.5%
    444
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.14
    0.01
    444
    +0.11
    0.01
    588
    +0.10
    0.01
    Negative Logits
     divor
    -0.60
    lishes
    -0.59
     legality
    -0.59
     weave
    -0.58
    arrell
    -0.58
    ilts
    -0.58
     tamp
    -0.57
    Recomm
    -0.56
     ribbon
    -0.56
    rious
    -0.56
    POSITIVE LOGITS
    ploy
    0.92
    selves
    0.90
    useum
    0.88
    ites
    0.85
    obile
    0.83
    aterial
    0.82
    asons
    0.81
    ayer
    0.77
    igrants
    0.77
    ograph
    0.76
    Activations Density 0.027%

    No Known Activations