INDEX

Explanations

phrases related to implying or suggesting something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "imply" or its variations

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

SEA

-0.77

Ĥ¬

-0.72

 Seymour

-0.69

qua

-0.69

Lago

-0.69

grass

-0.68

flix

-0.67

sth

-0.67

Total

-0.65

 Maid

-0.63

POSITIVE LOGITS

 impl

1.27

oded

1.11

anting

1.11

ausible

1.10

osion

1.07

icit

1.01

oding

0.94

astically

0.94

icating

0.93

anted

0.91

Activations Density 0.007%

phrases related to implying or suggesting something

instances of the word "imply" or its variations

No Comments

No Known Activations