INDEX

Explanations

mentions of variations of the word "provoke"

oai_token-act-pair · gpt-3.5-turbo

mentions of the city of Dnipro in Ukraine

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

cock

-0.78

ricks

-0.71

 cane

-0.70

 Grail

-0.70

geist

-0.69

hler

-0.69

 Scots

-0.67

horn

-0.64

erity

-0.64

hops

-0.64

POSITIVE LOGITS

ideo

0.98

incial

0.97

iral

0.95

irus

0.94

isions

0.92

itals

0.86

ially

0.86

ascular

0.84

nov

0.81

illard

0.81

Activations Density 0.034%

mentions of variations of the word "provoke"

mentions of the city of Dnipro in Ukraine

No Comments

No Known Activations