INDEX

Explanations

references to the name "Lance."

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Lance."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

worldly

-0.83

ocular

-0.79

icism

-0.77

sbm

-0.76

owl

-0.75

âĶģ

-0.75

alien

-0.75

ulators

-0.75

urated

-0.74

SpaceEngineers

-0.74

POSITIVE LOGITS

lot

1.30

 Corpor

1.01

 Armstrong

1.01

 Lance

0.90

 Daniels

0.82

 Redd

0.81

 Davidson

0.77

pool

0.77

 Stephenson

0.75

 Harvey

0.74

Activations Density 0.014%

references to the name "Lance."

mentions of the name "Lance."

No Comments

No Known Activations