© 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. 11-RES-JB
    11. 7143
    Prev
    Next
    INDEX
    Explanations

    verbs indicating the ability or inability of a subject to perform certain actions

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    reported
    -0.67
     Vest
    -0.65
    stories
    -0.65
    soDeliveryDate
    -0.64
     Lifetime
    -0.64
     Investig
    -0.64
    published
    -0.64
     pledges
    -0.63
     pledging
    -0.63
    accompanied
    -0.63
    POSITIVE LOGITS
     capitalize
    1.21
     dominate
    1.17
     score
    1.12
     overpower
    1.12
     penetrate
    1.11
     dictate
    1.10
     defend
    1.09
     overtake
    1.08
     execute
    1.08
     punish
    1.08
    Activations Density 0.116%

    No Known Activations