INDEX

Explanations

phrases indicating filtering or sorting actions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "by" indicating methods or means of action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

SPONSORED

-0.79

gow

-0.75

bard

-0.73

inea

-0.72

Reilly

-0.71

istar

-0.71

ETF

-0.69

aroo

-0.66

heimer

-0.66

ettes

-0.65

POSITIVE LOGITS

 virtue

1.24

products

1.11

 multiplying

0.96

 default

0.90

catch

0.86

product

0.86

 clicking

0.84

 adding

0.83

 removing

0.81

laws

0.81

Activations Density 0.159%

phrases indicating filtering or sorting actions

instances of the word "by" indicating methods or means of action

No Comments

No Known Activations