INDEX

Explanations

statements made by spokespersons or representatives

oai_token-act-pair · gpt-3.5-turbo

phrases indicating spokesperson statements

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

 ILCS

-0.94

à¦

-0.73

ï¸

-0.69

ãģ

-0.68

=~=~

-0.68

\/\/

-0.65

estern

-0.64

RIP

-0.63

ãģ®ç

-0.63

spir

-0.63

POSITIVE LOGITS

 briefs

0.74

terday

0.74

ogun

0.72

hement

0.70

 briefing

0.69

.):

0.69

arity

0.67

ardi

0.66

 briefed

0.65

 briefings

0.64

Activations Density 0.197%

statements made by spokespersons or representatives

phrases indicating spokesperson statements

No Comments

No Known Activations