INDEX

Explanations

phrases related to legal terms and criminal activities

oai_token-act-pair · gpt-3.5-turbo

words related to conversation or the act of convening

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

BALL

-0.77

HAM

-0.65

holders

-0.65

blind

-0.64

bledon

-0.63

WARE

-0.61

IGH

-0.60

THER

-0.60

erity

-0.60

 Pathfinder

-0.60

POSITIVE LOGITS

icts

1.41

oys

1.32

ocation

1.20

olution

1.18

ulsive

1.16

ulsion

1.10

iction

1.08

ocations

1.08

ince

1.06

ulse

1.06

Activations Density 0.022%

phrases related to legal terms and criminal activities

words related to conversation or the act of convening

No Comments

No Known Activations