INDEX

Explanations

words related to amendments or modifications

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "am" and its variants

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 aver

-0.77

 learners

-0.69

 takeaway

-0.67

 concession

-0.67

ttes

-0.66

mining

-0.63

 kinderg

-0.63

 Danger

-0.61

 Achievement

-0.61

DragonMagazine

-0.61

POSITIVE LOGITS

ethyst

1.27

sterdam

1.22

azon

1.17

ateurs

1.15

assing

1.10

endment

1.08

ajor

1.05

ilies

1.04

icable

1.02

essage

1.00

Activations Density 0.033%

words related to amendments or modifications

occurrences of the word "am" and its variants

No Comments

No Known Activations