© 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. 9-RES-JB
    11. 20053
    Prev
    Next
    INDEX
    Explanations

    recommendation phrases related to performing tasks or selecting options

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating certainty or determination about actions or outcomes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     pursuing
    -0.77
     withdrawing
    -0.76
     electing
    -0.71
     withholding
    -0.68
     wielding
    -0.67
    soever
    -0.66
     retreating
    -0.65
     negotiating
    -0.64
     Investig
    -0.64
     choosing
    -0.63
    POSITIVE LOGITS
     suffice
    1.49
     elicit
    1.06
     be
    1.04
     resonate
    1.03
     help
    1.00
     offend
    1.00
     confuse
    0.99
     attract
    0.98
     fit
    0.97
     annoy
    0.96
    Activations Density 0.327%

    No Known Activations