INDEX

Explanations

prominent names of individuals and characters in the text

oai_token-act-pair · gpt-3.5-turbo

references to notable individuals and their roles in significant events or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ãĤ´ãĥ³

-0.80

 prompting

-0.64

utterstock

-0.64

COMPLE

-0.60

azard

-0.60

Aug

-0.59

ãĤ¼

-0.58

BUT

-0.57

SEE

-0.57

 concurrent

-0.56

POSITIVE LOGITS

ain

1.49

 sucks

1.39

 shouldn

1.25

 deserves

1.23

 cannot

1.22

 doesn

1.20

 hasn

1.19

 hates

1.19

 doesnt

1.18

 gotta

1.16

Activations Density 0.718%

prominent names of individuals and characters in the text

references to notable individuals and their roles in significant events or statements

No Comments

No Known Activations