© 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. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 40470
    Prev
    Next
    INDEX
    Explanations

    names related to terrorism and Islamist themes

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    267
    +0.14
    0.6%
    526
    +0.13
    0.6%
    116
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.14
    0.19
    288
    +0.13
    0.18
    613
    +0.11
    0.08
    Negative Logits
     Occupations
    -0.72
     Clicker
    -0.65
     Morales
    -0.65
     Agents
    -0.62
     Consent
    -0.62
     Slave
    -0.62
    hani
    -0.60
    agents
    -0.60
    Gall
    -0.60
    ãģ¦
    -0.59
    POSITIVE LOGITS
    borgh
    1.22
    ming
    1.22
    ydia
    1.21
    ingo
    1.14
    orously
    1.14
    mer
    1.10
    beer
    1.09
    adesh
    1.06
    orous
    1.01
    inate
    1.00
    Activations Density 6.374%

    No Known Activations