© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 7190
    Prev
    Next
    INDEX
    Explanations

    phrases and constructions related to subjective opinions and personal beliefs

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    583
    +0.13
    0.6%
    158
    +0.10
    0.5%
    540
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.13
    0.10
    720
    +0.10
    0.09
    384
    +0.10
    0.05
    Negative Logits
    oother
    -0.85
    osate
    -0.75
    unic
    -0.75
    osponsors
    -0.75
     Rhod
    -0.74
     Buc
    -0.72
     Loren
    -0.70
    enium
    -0.70
    estern
    -0.70
    quished
    -0.69
    POSITIVE LOGITS
     actions
    1.39
     demeanor
    1.36
     feelings
    1.34
     behaviors
    1.32
     intentions
    1.32
     behavior
    1.30
     beliefs
    1.29
     behaviours
    1.28
     behaviour
    1.27
     attitudes
    1.25
    Activations Density 1.295%

    No Known Activations