INDEX

Explanations

names containing the letters "au."

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "au" or similar variations related to a specific context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

é¾į

-0.70

appers

-0.69

eyes

-0.67

 Cavaliers

-0.64

ors

-0.61

ALLY

-0.60

EVA

-0.59

 odds

-0.58

orative

-0.56

vic

-0.55

POSITIVE LOGITS

llah

1.30

gment

1.24

pload

1.14

fman

1.10

qua

1.09

clair

1.02

rosis

1.01

ction

1.00

lette

0.98

cham

0.95

Activations Density 0.042%

names containing the letters "au."

mentions of the word "au" or similar variations related to a specific context

No Comments

No Known Activations