INDEX

Explanations

noun forms and their derivatives related to abstract concepts and categories

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

phrases related to locations and positions within a setting

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Thing

-0.55

 Hunt

-0.52

 Rabbit

-0.49

Wem

-0.47

 Valkyrie

-0.46

anooga

-0.44

çĭ

-0.44

 Gunn

-0.43

Ku

-0.42

 Rapp

-0.41

POSITIVE LOGITS

terness

0.56

imity

0.55

theless

0.50

ishly

0.50

lly

0.46

terday

0.45

pedia

0.44

ysis

0.43

etheless

0.43

,[

0.42

Activations Density 0.558%

noun forms and their derivatives related to abstract concepts and categories

phrases related to locations and positions within a setting

No Comments

No Known Activations