© 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. 45997
    Prev
    Next
    INDEX
    Explanations

    instances of the word "probably" followed by an observation or statement

    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
    447
    +0.25
    1.0%
    718
    +0.13
    0.5%
    247
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.25
    0.15
    566
    +0.13
    0.10
    8
    +0.11
    0.11
    Negative Logits
    elight
    -0.86
    issy
    -0.84
    iya
    -0.84
    iates
    -0.83
    ife
    -0.81
    aily
    -0.80
    uctor
    -0.80
    iate
    -0.79
    church
    -0.76
    gian
    -0.76
    POSITIVE LOGITS
     misunder
    0.91
     quir
    0.83
     underestimate
    0.82
    Ń·
    0.82
     overest
    0.76
     exagger
    0.74
     owe
    0.74
     ali
    0.72
    TOR
    0.72
    ©¶æ
    0.72
    Activations Density 8.208%

    No Known Activations