© 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. 4202
    Prev
    Next
    INDEX
    Explanations

    references to legal issues, law enforcement activities, and criminal behavior

    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
    445
    +0.08
    0.5%
    481
    +0.08
    0.5%
    122
    +0.07
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.08
    0.26
    385
    +0.08
    0.23
    362
    +0.07
    0.19
    Negative Logits
    headers
    -0.60
    pointer
    -0.56
    peat
    -0.54
    pointers
    -0.53
     adapted
    -0.52
    intosh
    -0.52
    digy
    -0.51
     reprinted
    -0.50
     pitted
    -0.50
    hots
    -0.50
    POSITIVE LOGITS
    ilet
    0.59
     parliament
    0.54
    Ott
    0.50
     Fairfax
    0.50
     srfAttach
    0.49
     Parliament
    0.48
    asty
    0.47
    ichita
    0.47
    ggles
    0.46
     Senate
    0.45
    Activations Density 10.459%

    No Known Activations