INDEX

Explanations

prepositions indicating origin or distribution

oai_token-act-pair · gpt-3.5-turbo

references to numerical data and distribution patterns

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

assi

-0.76

iso

-0.74

äºĶ

-0.70

ilib

-0.68

©¶æ¥µ

-0.68

undo

-0.67

soType

-0.67

Ł

-0.67

DK

-0.66

dash

-0.66

POSITIVE LOGITS

 legitimate

0.83

 reputable

0.79

 residential

0.73

 actual

0.72

 genuine

0.71

 discrete

0.70

 lesser

0.69

 narrower

0.68

 poorer

0.68

 either

0.67

Activations Density 0.581%

prepositions indicating origin or distribution

references to numerical data and distribution patterns

No Comments

No Known Activations