INDEX

Explanations

capitalized strings denoting brand names or acronyms

oai_token-act-pair · gpt-3.5-turbo

instances of the abbreviation "RO" and its associations

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

flies

-0.94

ifications

-0.80

manship

-0.77

ãĥ¼ãĥĨãĤ£

-0.75

ussen

-0.74

riad

-0.71

hof

-0.70

urer

-0.69

geist

-0.69

hered

-0.69

POSITIVE LOGITS

OTS

1.28

JO

1.26

FILE

1.20

OT

1.16

KE

1.14

INT

1.13

OPER

1.11

LL

1.11

ILA

1.09

ADS

1.09

Activations Density 0.044%

capitalized strings denoting brand names or acronyms

instances of the abbreviation "RO" and its associations

No Comments

No Known Activations