INDEX

Explanations

instances of the word "visitor."

oai_token-act-pair · gpt-3.5-turbo

references to visitors and visitor experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

lain

-0.79

corn

-0.73

negie

-0.70

ray

-0.68

 marrow

-0.67

rum

-0.67

arity

-0.62

esson

-0.60

ells

-0.60

bra

-0.58

POSITIVE LOGITS

Reviewer

0.91

 attractions

0.88

 visitation

0.87

issance

0.86

hip

0.84

 visitors

0.82

cript

0.75

 attraction

0.74

glers

0.74

 visiting

0.74

Activations Density 0.018%

instances of the word "visitor."

references to visitors and visitor experiences

No Comments

No Known Activations