INDEX

Explanations

specific mentions of the letter "D" related to various contexts, such as names or abbreviations

oai_token-act-pair · gpt-3.5-turbo

references to the letter 'D' combined with numeric indicators

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

 Bulg

-0.80

ãĤ¡

-0.77

terday

-0.77

 Isles

-0.74

Nun

-0.66

 lett

-0.66

 Canaver

-0.66

 Remem

-0.65

 Ascension

-0.62

ierrez

-0.61

POSITIVE LOGITS

etermination

1.23

etermin

1.20

etermined

1.19

ownt

1.16

aughters

1.14

imensional

1.10

ynam

1.10

azz

1.10

izzy

1.08

urga

1.07

Activations Density 0.046%

specific mentions of the letter "D" related to various contexts, such as names or abbreviations

references to the letter 'D' combined with numeric indicators

No Comments

No Known Activations