INDEX

Explanations

references to the name "Obama"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "ama," potentially relating to specific places or names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ted

-0.98

ting

-0.94

tle

-0.86

lasses

-0.80

ritch

-0.77

points

-0.77

ttes

-0.77

lass

-0.75

teenth

-0.73

tto

-0.72

POSITIVE LOGITS

Ã±a

1.10

Ã±

1.00

ican

0.93

utra

0.92

udeau

0.92

qua

0.92

é¾įåĸļå£«

0.89

Ã®

0.82

eus

0.79

oka

0.79

Activations Density 0.063%

references to the name "Obama"

occurrences of the term "ama," potentially relating to specific places or names

No Comments

No Known Activations