INDEX

Explanations

mentions of the television show "Jeopardy"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Jeopardy."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ancial

-0.78

å§«

-0.77

raviolet

-0.75

IUM

-0.71

ãĥ´ãĤ¡

-0.70

ãĥ¼ãĥĨ

-0.69

ctica

-0.68

IFIED

-0.67

ional

-0.66

istics

-0.66

POSITIVE LOGITS

hovah

1.20

opard

1.17

zeb

1.00

pee

0.93

hov

0.92

erk

0.91

eps

0.90

orge

0.90

anty

0.87

apons

0.86

Activations Density 0.019%

mentions of the television show "Jeopardy"

occurrences of the word "Jeopardy."

No Comments

No Known Activations