INDEX

Explanations

references to Los Angeles, specifically with the abbreviation "LA"

oai_token-act-pair · gpt-3.5-turbo

mentions of the city "LA."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

lers

-0.84

lication

-0.80

ullah

-0.76

ãĥĥãĥĪ

-0.75

schild

-0.74

theless

-0.73

addons

-0.73

ãĥĩãĤ£

-0.72

ler

-0.71

manship

-0.71

POSITIVE LOGITS

UNCH

1.20

 Angeles

1.01

UGH

0.94

URA

0.91

KE

0.88

 Galaxy

0.83

 peninsula

0.78

 Lumpur

0.77

 Lakers

0.77

izabeth

0.74

Activations Density 0.008%

references to Los Angeles, specifically with the abbreviation "LA"

mentions of the city "LA."

No Comments

No Known Activations