INDEX

Explanations

system-related terms and technical instructions

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and special characters

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

 coup

-0.76

 ultr

-0.73

 Eston

-0.73

 dece

-0.72

 Semin

-0.71

 clut

-0.68

 Concord

-0.68

 Chamber

-0.68

 Ferr

-0.68

 fright

-0.68

POSITIVE LOGITS

meta

1.28

foo

1.26

true

1.19

insert

1.19

DATA

1.19

mega

1.18

input

1.15

SELECT

1.15

alpha

1.15

ctrl

1.14

Activations Density 0.103%

system-related terms and technical instructions

punctuation marks and special characters

No Comments

No Known Activations