Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 3-ATT-KK
    6. 10274
    Prev
    Next
    INDEX
    Explanations

    instances of the word "interesting" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_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.3.attn.hook_z
    Hook Layer
    3
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.02
    1:0.02
    2:0.06
    3:0.03
    4:0.02
    5:0.04
    6:0.06
    7:0.09
    8:0.04
    9:0.04
    10:0.10
    11:0.42
    Negative Logits
     canceled
    -1.22
    ���
    -1.19
    otent
    -1.19
    correct
    -1.15
    ドラ
    -1.11
     incorrectly
    -1.10
     cancel
    -1.08
     disqualified
    -1.07
     wrongly
    -1.05
     helicop
    -1.04
    POSITIVE LOGITS
     juxtap
    1.50
    yssey
    1.33
     contrasting
    1.31
     perspectives
    1.30
     insights
    1.26
     Alm
    1.21
     intrigue
    1.17
     newcom
    1.16
     twists
    1.15
     ⓘ
    1.14
    Activations Density 0.092%

    No Known Activations