© 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 125
    Prev
    Next
    INDEX
    Explanations

    mentions of spouses or partners

    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
    484
    +0.15
    0.5%
    235
    +0.14
    0.5%
    370
    +0.14
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    235
    +0.15
    0.05
    465
    +0.14
    0.04
    224
    +0.14
    0.04
    Negative Logits
    etting
    -1.24
     Flavoring
    -1.21
    oday
    -1.20
    umbn
    -1.09
    obyl
    -1.08
    anian
    -1.05
     GOODMAN
    -1.01
     constitu
    -1.01
    kefeller
    -0.99
    vernment
    -0.98
    POSITIVE LOGITS
    wife
    1.53
    hood
    1.25
    shake
    1.08
    parts
    1.05
    killer
    1.04
    fol
    1.02
    lessness
    1.01
    mate
    1.00
    maker
    1.00
    mother
    0.99
    Activations Density 0.584%

    No Known Activations