© 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. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 9-ATT-KK
    11. 22616
    Prev
    Next
    INDEX
    Explanations

    phrases related to speculation, misinformation, and claims of false reporting

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L9_Hcat_z_lr1.20e-03_l11.20e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.9.attn.hook_z
    Hook Layer
    9
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.09
    1:0.01
    2:0.08
    3:0.41
    4:0.03
    5:0.09
    6:0.01
    7:0.05
    8:0.03
    9:0.01
    10:0.12
    11:0.02
    Negative Logits
    ridor
    -2.32
     reunited
    -2.21
     stationed
    -2.18
     Nanto
    -2.11
     retiring
    -2.10
     renov
    -2.01
    externalActionCode
    -2.01
    pora
    -2.00
    conservancy
    -1.99
     Shrine
    -1.97
    POSITIVE LOGITS
     falsehood
    3.78
     misinformation
    3.73
     nonsense
    3.67
     inaccurate
    3.59
     inaccur
    3.53
     facts
    3.49
     myths
    3.48
     baseless
    3.45
     explanations
    3.44
     bullshit
    3.44
    Activations Density 1.496%

    No Known Activations