INDEX

Explanations

mentions of the name "Lore" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to a person named Loretta or related lore

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

akra

-0.80

obs

-0.71

urers

-0.69

ixties

-0.67

etheus

-0.66

inished

-0.65

insula

-0.65

onement

-0.64

undred

-0.63

isive

-0.62

POSITIVE LOGITS

 Lore

1.13

nz

1.07

tta

1.01

walker

0.90

lette

0.90

gon

0.85

gate

0.84

tto

0.84

bolt

0.82

nce

0.81

Activations Density 0.013%

mentions of the name "Lore" in various contexts

references to a person named Loretta or related lore

No Comments

No Known Activations