© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 3-RES-JB
    11. 16726
    Prev
    Next
    INDEX
    Explanations

    words related to manipulation or manipulative actions

    oai_token-act-pair · gpt-3.5-turbo

    terms related to manipulation and control

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çĦ
    -0.76
     Blessed
    -0.74
    Republic
    -0.71
    hood
    -0.70
     Glory
    -0.70
    gap
    -0.69
    ness
    -0.68
    zl
    -0.68
     Cic
    -0.68
     Fallen
    -0.68
    POSITIVE LOGITS
     manipulate
    1.14
     manip
    1.11
     manipulating
    1.02
     manipulation
    0.97
     manipulated
    0.96
    glers
    0.86
     bait
    0.83
    ationally
    0.82
     barg
    0.81
     levers
    0.80
    Activations Density 0.012%

    No Known Activations