© 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. 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. 14164
    Prev
    Next
    INDEX
    Explanations

    the word "other"

    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
    393
    +0.21
    0.7%
    159
    +0.17
    0.6%
    101
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    159
    +0.21
    0.05
    402
    +0.17
    0.05
    621
    +0.14
    0.05
    Negative Logits
     Retrieved
    -0.88
    ulas
    -0.83
    gets
    -0.82
    yip
    -0.79
    ulner
    -0.78
    Lee
    -0.77
     Patriarch
    -0.77
     Ambassador
    -0.77
     Trophy
    -0.77
    Always
    -0.76
    POSITIVE LOGITS
    worldly
    2.09
    wise
    1.27
    words
    1.24
    world
    1.13
     respects
    1.10
    word
    1.08
     ways
    1.02
     vein
    0.97
     words
    0.95
     manner
    0.94
    Activations Density 0.373%

    No Known Activations