© 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. 2-RES-JB
    11. 13379
    Prev
    Next
    INDEX
    Explanations

    words related to legal or criminal activities, such as 'suspect', 'officer', 'robber', and 'victim'

    oai_token-act-pair · gpt-3.5-turbo

    references to suspects and police interactions related to criminal incidents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Flavoring
    -0.67
    okers
    -0.63
    laus
    -0.61
    BN
    -0.61
    icion
    -0.61
    imum
    -0.60
    ysical
    -0.60
    rising
    -0.60
    emonic
    -0.60
    annabin
    -0.59
    POSITIVE LOGITS
     assailant
    0.70
     rapist
    0.68
     Giul
    0.67
     examiner
    0.66
     sergeant
    0.65
    hood
    0.64
     pleaded
    0.63
    """
    0.63
     Theodore
    0.62
     Romeo
    0.62
    Activations Density 0.407%

    No Known Activations