INDEX

Explanations

names or variations of the name "Sher" occurring in the text

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Sher."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Transform

-0.75

 projector

-0.72

ko

-0.71

iso

-0.70

 cycle

-0.68

 eggs

-0.67

 transformer

-0.66

gg

-0.66

 degrade

-0.66

abyte

-0.65

POSITIVE LOGITS

 Sher

3.89

Sher

2.69

 sher

2.31

 Sherman

1.42

 Cher

1.25

 Sheridan

1.22

Kyr

1.19

 Sheriff

1.19

 Christy

1.14

 Sherlock

1.13

Activations Density 0.012%

names or variations of the name "Sher" occurring in the text

mentions of the name "Sher."

No Comments

No Known Activations