© 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. 34834
    Prev
    Next
    INDEX
    Explanations

    commands and code snippets related to software development and system administration

    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
    141
    +0.08
    0.6%
    61
    +0.07
    0.5%
    385
    +0.06
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    294
    +0.08
    0.03
    640
    +0.07
    0.24
    314
    +0.06
    -0.10
    Negative Logits
    iasis
    -0.47
    acci
    -0.46
     induct
    -0.42
     streaks
    -0.42
     waves
    -0.42
    holes
    -0.41
    aciously
    -0.40
    abus
    -0.40
    osion
    -0.40
     omission
    -0.39
    POSITIVE LOGITS
    \
    0.56
    SpaceEngineers
    0.53
    \/
    0.51
    \<
    0.51
    ãĥĦ
    0.51
    ~
    0.51
    bryce
    0.50
    username
    0.50
    /$
    0.49
     ~/.
    0.48
    Activations Density 13.452%

    No Known Activations