INDEX

Explanations

references to the concept of "man," often referring to males or masculinity

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "man."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Grounds

-0.71

 Ames

-0.67

ADC

-0.63

AVG

-0.63

 Oath

-0.61

DERR

-0.60

 Rivals

-0.59

 outgoing

-0.59

 Dame

-0.59

viz

-0.58

POSITIVE LOGITS

gling

1.33

hattan

1.32

ifest

1.30

nered

1.25

uscript

1.19

oeuv

1.16

agers

1.16

handled

1.14

made

1.13

ila

1.12

Activations Density 0.059%

references to the concept of "man," often referring to males or masculinity

occurrences of the word "man."

No Comments

No Known Activations