INDEX

Explanations

verbs related to actions or events

oai_token-act-pair · gpt-3.5-turbo

statements about deception or manipulation

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

accompanied

-0.75

fter

-0.68

fters

-0.65

critical

-0.65

ggles

-0.63

foreseen

-0.62

 Approximately

-0.62

è¦ļéĨĴ

-0.60

uers

-0.60

cone

-0.60

POSITIVE LOGITS

 themselves

1.39

 THEIR

0.88

 their

0.85

us

0.81

li

0.80

 uniforms

0.72

selves

0.71

 fools

0.70

 selves

0.66

me

0.65

Activations Density 0.714%

verbs related to actions or events

statements about deception or manipulation

No Comments

No Known Activations