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. 5-ATT-KK
    6. 43923
    Prev
    Next
    INDEX
    Explanations

    statements or phrases involving denial and conflicting claims

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    Data Type
    float32
    Hook Name
    blocks.5.attn.hook_z
    Hook Layer
    5
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.06
    1:0.02
    2:0.04
    3:0.11
    4:0.03
    5:0.16
    6:0.01
    7:0.02
    8:0.06
    9:0.24
    10:0.13
    11:0.07
    Negative Logits
    ortmund
    -1.07
    ortium
    -1.03
    racuse
    -0.93
    DragonMagazine
    -0.93
    untled
    -0.92
    hedral
    -0.91
    RPG
    -0.88
    marked
    -0.88
    ¶
    -0.88
    pletion
    -0.87
    POSITIVE LOGITS
    .'"
    1.83
    ..."
    1.76
     …"
    1.66
    )."
    1.59
    '."
    1.56
    !'"
    1.51
    .""
    1.49
    ]."
    1.45
    ?'"
    1.45
    .")
    1.44
    Activations Density 2.508%

    No Known Activations