INDEX

Explanations

mentions of a particular acronym or abbreviation, specifically "SP"

oai_token-act-pair · gpt-3.5-turbo

the presence of the abbreviation "SP" or specific codes related to a structured format or categorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

taboola

-0.77

sbm

-0.72

ãĥĥãĥĪ

-0.71

 behold

-0.70

 subord

-0.66

 chronological

-0.64

 journalistic

-0.63

omorph

-0.62

naires

-0.61

 Exodus

-0.60

POSITIVE LOGITS

EED

1.40

OIL

1.29

ORTS

1.27

ORT

1.20

ECT

1.07

ONSORED

1.03

IR

1.01

ERSON

0.99

orter

0.98

ENCE

0.97

Activations Density 0.014%

mentions of a particular acronym or abbreviation, specifically "SP"

the presence of the abbreviation "SP" or specific codes related to a structured format or categorization

No Comments

No Known Activations