INDEX

Explanations

words related to manipulation or sales techniques

oai_token-act-pair · gpt-3.5-turbo

words associated with textures and conditions of items or situations

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

terday

-0.78

USE

-0.66

 observation

-0.60

 lapse

-0.60

 vision

-0.60

 Copenhagen

-0.60

VIS

-0.59

é¾įå¥ĳå£«

-0.59

ISM

-0.58

ĨĴ

-0.58

POSITIVE LOGITS

led

1.52

les

1.42

ling

1.37

lers

1.33

ler

1.19

lies

1.04

le

1.02

eling

1.02

lin

1.01

lement

0.98

Activations Density 0.040%

words related to manipulation or sales techniques

words associated with textures and conditions of items or situations

No Comments

No Known Activations