INDEX

Explanations

common acronyms or proper nouns in a structured format

oai_token-act-pair · gpt-3.5-turbo

instances of the term "OT" or related abbreviations in various contexts

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

antha

-0.71

iasis

-0.71

ously

-0.70

oused

-0.68

bler

-0.65

perate

-0.60

ity

-0.58

 equation

-0.58

icides

-0.58

 reasoned

-0.58

POSITIVE LOGITS

assium

1.06

TL

1.03

ALLY

0.93

TO

0.92

TE

0.92

TI

0.88

tery

0.86

atoes

0.81

OGR

0.80

ION

0.79

Activations Density 0.035%

common acronyms or proper nouns in a structured format

instances of the term "OT" or related abbreviations in various contexts

No Comments

No Known Activations