INDEX

Explanations

proper nouns or names consisting of two parts

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "eb" in various contexts

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

ivities

-0.77

 nomine

-0.73

FTWARE

-0.70

ivity

-0.68

 skirts

-0.66

 butterflies

-0.65

 Inquisitor

-0.64

 Ashe

-0.64

folk

-0.63

 pole

-0.61

POSITIVE LOGITS

ruary

1.41

oard

1.11

edia

1.03

acteria

1.00

ulous

0.98

uilt

0.98

rahim

0.97

rief

0.94

bies

0.92

uild

0.90

Activations Density 0.015%

proper nouns or names consisting of two parts

the presence of the word "eb" in various contexts

No Comments

No Known Activations