© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 40931
    Prev
    Next
    INDEX
    Explanations

    phrases related to the concept of going "too far" in various contexts

    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
    285
    +0.14
    0.5%
    726
    +0.13
    0.5%
    55
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.14
    0.21
    636
    +0.13
    0.07
    567
    +0.13
    0.08
    Negative Logits
    hyde
    -0.98
    quist
    -0.85
    idia
    -0.81
    Script
    -0.81
    issance
    -0.78
    inarily
    -0.78
    creator
    -0.77
    payer
    -0.75
    riots
    -0.75
     inventor
    -0.74
    POSITIVE LOGITS
    oths
    1.02
     risky
    0.92
     much
    0.90
     stren
    0.89
     lax
    0.87
    avorite
    0.87
     far
    0.87
     BAD
    0.87
     tempting
    0.86
    apo
    0.84
    Activations Density 10.294%

    No Known Activations