INDEX

Explanations

phrases related to lists or items that are being grouped together under a common category

oai_token-act-pair · gpt-3.5-turbo

conjunctions and transitional phrases that connect ideas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

abul

-0.69

scrib

-0.68

athetic

-0.67

lear

-0.66

busters

-0.65

natureconservancy

-0.64

habi

-0.64

tera

-0.63

onomic

-0.62

spir

-0.61

POSITIVE LOGITS

 finally

1.68

 then

1.16

vo

1.07

 Finally

1.06

 Lastly

1.04

 thence

1.04

 THEN

1.02

 prest

0.98

 assorted

0.97

Lastly

0.95

Activations Density 0.206%

phrases related to lists or items that are being grouped together under a common category

conjunctions and transitional phrases that connect ideas

No Comments

No Known Activations