INDEX

Explanations

words related to a specific proper noun, "Greenspan"

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific entity or term related to "pan."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 GEAR

-0.73

VERTIS

-0.71

ITNESS

-0.70

 FANTASY

-0.68

TOD

-0.67

çĦ

-0.67

VEL

-0.67

é¾

-0.65

DIR

-0.65

DEF

-0.64

POSITIVE LOGITS

pan

1.49

zees

0.98

olin

0.91

Pan

0.86

obal

0.84

asonic

0.83

ning

0.83

opol

0.81

eful

0.81

igans

0.80

Activations Density 0.004%

words related to a specific proper noun, "Greenspan"

mentions of a specific entity or term related to "pan."

No Comments

No Known Activations