INDEX

Explanations

words related to people's names or roles

oai_token-act-pair · gpt-3.5-turbo

the preposition "in"

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

 Seym

-0.90

 redund

-0.75

 destro

-0.73

 warr

-0.72

 millenn

-0.69

catentry

-0.69

hovah

-0.69

¶æ

-0.68

 payday

-0.68

 reluct

-0.67

POSITIVE LOGITS

strument

1.44

aug

1.24

jury

1.18

iti

1.16

ners

1.13

flation

1.12

jured

1.11

hibited

1.10

cluded

1.08

ned

1.07

Activations Density 0.051%

words related to people's names or roles

the preposition "in"

No Comments

No Known Activations