INDEX

Explanations

commands or directives given in text

oai_token-act-pair · gpt-3.5-turbo

directives or calls to action in the text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sacrific

-0.73

)].

-0.66

 mono

-0.66

 decoration

-0.65

 Haku

-0.65

 holster

-0.65

 binding

-0.64

UD

-0.62

 bottleneck

-0.61

-"

-0.61

POSITIVE LOGITS

resa

1.05

anamo

0.92

pedia

0.90

News

0.83

lahoma

0.83

utterstock

0.82

chin

0.81

akov

0.81

htaking

0.80

azeera

0.80

Activations Density 0.455%

commands or directives given in text

directives or calls to action in the text

No Comments

No Known Activations