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. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 854
    Prev
    Next
    INDEX
    Explanations

    the verb "launch" followed by a specific target or action

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    288
    +0.28
    0.9%
    665
    +0.17
    0.5%
    471
    +0.17
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.28
    0.07
    541
    +0.17
    0.05
    471
    +0.17
    0.04
    Negative Logits
    GO
    -1.07
    iane
    -1.02
    nton
    -1.02
     Marketable
    -1.02
    Son
    -0.99
    Wa
    -0.97
    uth
    -0.97
    ipped
    -0.95
    uala
    -0.93
    iddler
    -0.93
    POSITIVE LOGITS
    pad
    1.60
    launch
    1.40
    eers
    1.38
    nings
    1.25
    laun
    1.21
    lishing
    1.21
    lift
    1.19
     launching
    1.19
    eer
    1.18
    escape
    1.15
    Activations Density 1.011%

    No Known Activations