© 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 EvalsExportsAPI 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. 13046
    Prev
    Next
    INDEX
    Explanations

    words related to mistakes or errors

    oai_token-act-pair · gpt-3.5-turbo

    words or phrases that indicate mistakes or mismanagement

    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
    levels
    -0.81
    fast
    -0.76
    boat
    -0.73
    join
    -0.73
    thood
    -0.72
    ovy
    -0.70
    ready
    -0.70
    hene
    -0.67
    acl
    -0.67
    allic
    -0.66
    POSITIVE LOGITS
     mis
    3.30
     misinterpret
    2.06
     misunderstand
    1.92
    mis
    1.86
    Mis
    1.85
     misrepresent
    1.80
     Mis
    1.79
     mistaken
    1.73
     mishand
    1.73
     misunderstanding
    1.72
    Activations Density 0.026%

    No Known Activations