INDEX

Explanations

proper nouns related to historical figures or places

oai_token-act-pair · gpt-3.5-turbo

references to historical figures, specifically those named Ferdinand

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SF

-0.72

odi

-0.72

 Apex

-0.72

fters

-0.71

atham

-0.70

eah

-0.68

alon

-0.68

 Kend

-0.68

NJ

-0.67

Tree

-0.66

POSITIVE LOGITS

 Ferdinand

3.52

Ð´

2.31

dinand

1.71

ienne

1.12

 deposition

1.01

ï¿½ï¿½

1.01

ï¿½ï¿½ï¿½ï¿½

0.97

 Gupta

0.96

Fer

0.96

ocide

0.96

Activations Density 0.052%

proper nouns related to historical figures or places

references to historical figures, specifically those named Ferdinand

No Comments

No Known Activations