INDEX

Explanations

mentions of individuals named "Am" or "Ama"

oai_token-act-pair · gpt-3.5-turbo

the repeated mention of the name "Am."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ãģį

-0.70

 wart

-0.70

 directions

-0.69

 concession

-0.69

 diving

-0.69

 aver

-0.68

EStreamFrame

-0.65

 learners

-0.63

theless

-0.62

vous

-0.62

POSITIVE LOGITS

ethyst

1.32

bitious

1.29

sterdam

1.23

endment

1.19

ulet

1.18

nesia

1.14

nesty

1.11

ateur

1.03

icus

1.02

ajor

1.01

Activations Density 0.017%

mentions of individuals named "Am" or "Ama"

the repeated mention of the name "Am."

No Comments

No Known Activations