© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 2844
    Prev
    Next
    INDEX
    Explanations

    terms related to undercover operations

    oai_token-act-pair · gpt-3.5-turbo

    mentions of undercover operations or personnel

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    KB
    -0.66
     Races
    -0.66
    ament
    -0.64
    rist
    -0.63
    lag
    -0.63
     Pwr
    -0.62
    MB
    -0.62
     heap
    -0.61
     Rel
    -0.60
     Fract
    -0.60
    POSITIVE LOGITS
     undercover
    3.76
     underage
    1.30
     informant
    1.26
     infiltr
    1.23
     informants
    1.19
     investigative
    1.15
     unwitting
    1.08
     covert
    1.06
     clandestine
    1.05
     disguised
    1.04
    Activations Density 0.014%

    No Known Activations