INDEX

Explanations

initials or acronyms related to organizations or groups

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms commonly used in a specific context, likely related to organizations or technical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

-0.73

ãĥĸ

-0.68

alam

-0.64

 Sakuya

-0.64

 compan

-0.63

ru

-0.61

ãĤ¨ãĥ«

-0.60

Ã¢

-0.60

bern

-0.60

 func

-0.59

POSITIVE LOGITS

HY

0.81

ITION

0.80

KA

0.79

senal

0.79

HS

0.78

KI

0.76

ESH

0.75

NT

0.75

ZA

0.74

INESS

0.74

Activations Density 0.117%

initials or acronyms related to organizations or groups

abbreviations or acronyms commonly used in a specific context, likely related to organizations or technical terms

No Comments

No Known Activations