INDEX

Explanations

mentions of the word 'att' followed by a number

oai_token-act-pair · gpt-3.5-turbo

variations of the word "att."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 jaws

-0.67

 fentanyl

-0.64

phe

-0.63

 Uzbek

-0.62

 Galile

-0.62

vous

-0.60

 methyl

-0.60

Ö¼

-0.59

 Commando

-0.59

lda

-0.59

POSITIVE LOGITS

anooga

1.54

leground

1.27

ributes

1.21

itude

1.19

ributed

1.13

hews

1.13

orney

1.12

ista

1.10

erers

1.10

erer

1.08

Activations Density 0.041%

mentions of the word 'att' followed by a number

variations of the word "att."

No Comments

No Known Activations