INDEX

Explanations

phrases indicating the addition of information or commentary

oai_token-act-pair · gpt-3.5-turbo

instances of the word "adds," indicating the addition of information or opinions 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 @ 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

AH

-0.76

ograms

-0.73

view

-0.69

 Antar

-0.65

 Gram

-0.65

FL

-0.64

bia

-0.63

bin

-0.62

 Behavioral

-0.61

Fla

-0.61

POSITIVE LOGITS

endum

1.05

itional

0.91

itionally

0.90

 ingred

0.90

Adds

0.89

 challeng

0.84

eston

0.83

 insult

0.82

resso

0.82

accompan

0.81

Activations Density 0.004%

phrases indicating the addition of information or commentary

instances of the word "adds," indicating the addition of information or opinions in various contexts

No Comments

No Known Activations