INDEX

Explanations

proper nouns, specifically the name "Lindsey Graham"

oai_token-act-pair · gpt-3.5-turbo

the name "Lindsey Graham."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

arist

-0.83

haps

-0.79

avez

-0.71

awaru

-0.70

ibli

-0.70

ropri

-0.68

requisite

-0.68

asta

-0.68

rahim

-0.67

yright

-0.67

POSITIVE LOGITS

 Lindsey

1.20

 Buckingham

0.96

 Graham

0.92

 Hendricks

0.84

say

0.84

 Stephens

0.81

wark

0.80

 Palmer

0.79

Von

0.79

 Stewart

0.78

Activations Density 0.006%

proper nouns, specifically the name "Lindsey Graham"

the name "Lindsey Graham."

No Comments

No Known Activations