INDEX

Explanations

phrases or words related to referencing or mentioning specific information

oai_token-act-pair · gpt-3.5-turbo

repeated references to mentioned concepts or topics within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

isol

-0.76

apo

-0.68

ffic

-0.66

PU

-0.65

capt

-0.64

bor

-0.64

iolet

-0.63

 Clicker

-0.62

idity

-0.62

cession

-0.61

POSITIVE LOGITS

 herein

0.92

 above

0.83

 Parenthood

0.81

 inconsist

0.78

by

0.76

 hereafter

0.75

 è£ıè

0.75

 therein

0.74

 supra

0.74

ãĤ´ãĥ³

0.71

Activations Density 0.180%

phrases or words related to referencing or mentioning specific information

repeated references to mentioned concepts or topics within the text

No Comments

No Known Activations