INDEX

Explanations

mentions of objects or actions that involve automation or automatic processes

oai_token-act-pair · gpt-3.5-turbo

instances of the word "automatic" 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

borough

-0.92

ĸļ

-0.82

apest

-0.82

our

-0.76

usky

-0.75

rug

-0.75

orks

-0.74

Else

-0.74

nard

-0.71

IVERS

-0.71

POSITIVE LOGITS

 transmission

1.22

 renewal

0.95

 transmissions

0.95

 braking

0.94

omatic

0.90

 induction

0.89

 automatic

0.89

 confir

0.88

 termination

0.87

 captcha

0.86

Activations Density 0.019%

mentions of objects or actions that involve automation or automatic processes

instances of the word "automatic" in various contexts

No Comments

No Known Activations