INDEX

Explanations

pronouns followed by actions related to physical interactions or behaviors involving another person

oai_token-act-pair · gpt-3.5-turbo

pronouns and references to a male subject

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Bundes

-0.74

nb

-0.72

Mae

-0.70

carb

-0.65

 Marketable

-0.65

Counter

-0.64

utical

-0.63

counter

-0.63

CNN

-0.61

Fuel

-0.60

POSITIVE LOGITS

ctor

0.80

 semen

0.79

ading

0.73

sing

0.72

 ejac

0.71

 penis

0.71

 handsome

0.70

 sperm

0.69

avier

0.68

 circumcised

0.67

Activations Density 0.299%

pronouns followed by actions related to physical interactions or behaviors involving another person

pronouns and references to a male subject

No Comments

No Known Activations