© 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. 7261
    Prev
    Next
    INDEX
    Explanations

    negative evaluations of actions and behaviors

    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
    285
    +0.13
    0.6%
    29
    +0.11
    0.5%
    683
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    285
    +0.13
    0.05
    197
    +0.11
    0.06
    762
    +0.10
    0.08
    Negative Logits
     slips
    -0.67
     sightings
    -0.66
     stocks
    -0.61
    atonin
    -0.61
    ellar
    -0.60
     mutated
    -0.60
    anned
    -0.59
    amous
    -0.58
    selected
    -0.58
     dominate
    -0.58
    POSITIVE LOGITS
    nesty
    0.74
     hypocritical
    0.69
     disrespectful
    0.67
     unworthy
    0.66
     irresponsible
    0.65
    ãħĭ
    0.65
     excuse
    0.65
     insulting
    0.64
    åī
    0.64
    amount
    0.64
    Activations Density 1.170%

    No Known Activations