INDEX

Explanations

phrases related to sneaky or deceitful behavior

oai_token-act-pair · gpt-3.5-turbo

words containing the syllable "sle" or related to laziness and sloppiness

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

interrupted

-0.82

cake

-0.79

eous

-0.73

uates

-0.73

SHIP

-0.72

iates

-0.70

ipolar

-0.69

oulos

-0.67

IO

-0.66

iating

-0.65

POSITIVE LOGITS

azy

1.06

igh

0.99

uth

0.90

eker

0.89

sle

0.89

asure

0.89

pload

0.86

vered

0.85

ight

0.80

combe

0.78

Activations Density 0.015%

phrases related to sneaky or deceitful behavior

words containing the syllable "sle" or related to laziness and sloppiness

No Comments

No Known Activations