INDEX

Explanations

euphemisms used in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to euphemisms and landscaping

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

hold

-0.76

ly

-0.75

hart

-0.70

çīĪ

-0.69

lings

-0.69

 Grind

-0.69

 Protector

-0.67

watch

-0.67

rah

-0.66

weight

-0.66

POSITIVE LOGITS

isman

0.81

okin

0.81

istical

0.81

veyard

0.81

arial

0.80

istically

0.80

andum

0.79

ename

0.77

apons

0.77

apers

0.76

Activations Density 0.019%

euphemisms used in various contexts

terms related to euphemisms and landscaping

No Comments

No Known Activations