INDEX

Explanations

pronouns followed by positive or admiring sentiments

oai_token-act-pair · gpt-3.5-turbo

instances of the word "It" highlighting the importance of certain statements or observations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Eighth

-0.68

warning

-0.61

itatively

-0.60

anton

-0.60

 è£ıè¦ļéĨĴ

-0.57

arthed

-0.57

NX

-0.56

 Mole

-0.56

 Dayton

-0.56

 Mouth

-0.55

POSITIVE LOGITS

ain

1.15

chy

1.11

 hurts

1.05

zbollah

1.05

 wasn

1.04

 seems

1.02

 happened

1.02

 happens

1.00

 depends

0.97

 boils

0.96

Activations Density 0.266%

pronouns followed by positive or admiring sentiments

instances of the word "It" highlighting the importance of certain statements or observations

No Comments

No Known Activations