INDEX

Explanations

phrases related to a variety of topics or topics with multiple elements

oai_token-act-pair · gpt-3.5-turbo

quantifiers and descriptors related to various categories or types of entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ften

-0.75

enough

-0.71

answer

-0.71

aucus

-0.70

Enough

-0.66

ileaks

-0.65

Answer

-0.64

appointed

-0.64

actionDate

-0.63

/>

-0.63

POSITIVE LOGITS

 types

1.52

 kinds

1.49

 aspects

1.38

 forms

1.24

 industries

1.16

 facets

1.15

 sectors

1.10

 sorts

1.08

 categories

1.00

 areas

0.97

Activations Density 0.261%

phrases related to a variety of topics or topics with multiple elements

quantifiers and descriptors related to various categories or types of entities

No Comments

No Known Activations