© 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
    2. Under Peer Review · Attention SAE Research Paper
    3. GPT2-Small
    4. Attention Out
    5. 2-ATT-KK
    6. 22769
    Prev
    Next
    INDEX
    Explanations

    words related to technical information or items, potentially related to computers or technology

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    words associated with actions, states of being, and attributes of objects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L2_Hcat_z_lr1.20e-03_l11.00e+00_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v4.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.2.attn.hook_z
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.07
    1:0.05
    2:0.23
    3:0.11
    4:0.04
    5:0.07
    6:0.04
    7:0.02
    8:0.11
    9:0.10
    10:0.08
    11:0.03
    Negative Logits
    iots
    -1.32
    racial
    -1.28
    attacks
    -1.28
    humans
    -1.26
     sucked
    -1.25
    essor
    -1.25
     blues
    -1.24
     sucks
    -1.23
     scares
    -1.22
     blows
    -1.21
    POSITIVE LOGITS
    ��
    1.64
    ��極
    1.58
     CNBC
    1.35
    CLUD
    1.34
    ��
    1.31
    GMT
    1.29
    �
    1.28
     Hermes
    1.26
    ゴ
    1.25
    ��
    1.24
    Activations Density 0.134%

    No Known Activations