INDEX

Explanations

words related to locations or numbers

oai_token-act-pair · gpt-3.5-turbo

the verb form 'ied' 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 @ 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

arta

-0.90

OUNT

-0.86

Effective

-0.76

ivia

-0.74

Solution

-0.73

wcsstore

-0.72

Reviewer

-0.71

ajo

-0.70

izo

-0.69

ities

-0.68

POSITIVE LOGITS

vironment

0.79

utenant

0.78

eer

0.72

ppelin

0.72

 Machina

0.71

doms

0.69

leck

0.68

ept

0.67

ezvous

0.67

vous

0.67

Activations Density 0.025%

words related to locations or numbers

the verb form 'ied' in various contexts

No Comments

No Known Activations