© 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. 0-RES-JB
    11. 11416
    Prev
    Next
    INDEX
    Explanations

    phrases related to poor performance or quality

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "poorly" and phrases indicating poor performance or quality

    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
    aura
    -0.64
    Joy
    -0.64
     Purs
    -0.61
    tu
    -0.59
     Proof
    -0.59
     Lift
    -0.59
    au
    -0.59
    ity
    -0.58
    esse
    -0.58
     extradition
    -0.58
    POSITIVE LOGITS
     poorly
    3.51
     badly
    2.07
     incorrectly
    1.65
     inadequ
    1.59
     inconsist
    1.58
     horribly
    1.58
     improperly
    1.58
     inefficient
    1.42
    poor
    1.41
     lousy
    1.39
    Activations Density 0.017%

    No Known Activations