© 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. 45747
    Prev
    Next
    INDEX
    Explanations

    technical terms or actions related to bypassing or overriding certain systems or processes

    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
    +0.19
    0.9%
    218
    +0.10
    0.5%
    622
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.19
    0.10
    95
    +0.10
    0.04
    214
    +0.10
    0.04
    Negative Logits
    ankind
    -0.69
    Kind
    -0.67
    cas
    -0.66
    soType
    -0.61
    ãĤ·ãĥ£
    -0.61
     breath
    -0.60
    akh
    -0.60
    Baby
    -0.60
     Gael
    -0.60
    NAS
    -0.59
    POSITIVE LOGITS
    ricular
    0.89
    ibly
    0.84
     bypass
    0.83
    ed
    0.80
    es
    0.72
    edIn
    0.71
    ibility
    0.71
    ing
    0.69
    yz
    0.69
    ython
    0.69
    Activations Density 10.132%

    No Known Activations