INDEX

Explanations

words related to physical space or size, such as measurements and dimensions

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "footprint" 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

icum

-0.77

itals

-0.71

erest

-0.70

actionDate

-0.68

regulated

-0.67

rators

-0.66

ending

-0.66

ifying

-0.65

sell

-0.65

kaya

-0.65

POSITIVE LOGITS

 footprint

1.24

 footprints

1.08

////////////////

0.73

print

0.71

 Shoals

0.70

 Trash

0.67

eless

0.64

ystem

0.64

gie

0.64

=~

0.63

Activations Density 0.017%

words related to physical space or size, such as measurements and dimensions

references to the concept of "footprint" in various contexts

No Comments

No Known Activations