INDEX

Explanations

references to the word "coins" at different activation levels

oai_token-act-pair · gpt-3.5-turbo

references to coins

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Western

-0.71

Jol

-0.71

NEY

-0.69

ROR

-0.68

ECH

-0.67

 Lynd

-0.64

 Narr

-0.64

 Loud

-0.63

Neigh

-0.63

 Doctrine

-0.63

POSITIVE LOGITS

 coins

1.30

coins

1.13

urance

1.05

mith

0.99

folios

0.97

ilver

0.94

 coin

0.93

 mint

0.92

rencies

0.91

avorite

0.90

Activations Density 0.007%

references to the word "coins" at different activation levels

references to coins

No Comments

No Known Activations