© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4792
    Prev
    Next
    INDEX
    Explanations

    terms related to undermining or damaging something, particularly in a metaphorical or societal context

    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%
    497
    +0.12
    0.5%
    3
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    3
    +0.13
    0.04
    288
    +0.12
    0.04
    285
    +0.10
    0.02
    Negative Logits
     evict
    -0.65
    odus
    -0.65
    lov
    -0.61
    CENT
    -0.61
    eez
    -0.60
    sky
    -0.58
    ILLE
    -0.57
    çĦ
    -0.57
    ãĥ£
    -0.57
    hops
    -0.56
    POSITIVE LOGITS
    ments
    1.01
    ment
    0.85
    mental
    0.81
     undermin
    0.79
    luster
    0.79
     perceptions
    0.75
     confidence
    0.73
     susceptibility
    0.72
    ingly
    0.70
    manship
    0.69
    Activations Density 0.210%

    No Known Activations