INDEX

Explanations

proper nouns, specifically the name "Marc"

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

FUL

-0.78

itness

-0.78

OTE

-0.72

LLOW

-0.71

 SEAL

-0.70

worldly

-0.68

velt

-0.67

irlf

-0.65

ga

-0.64

 habitable

-0.64

POSITIVE LOGITS

agne

1.02

afort

0.97

ouf

0.94

angelo

0.90

ques

0.89

anson

0.88

iano

0.86

ias

0.85

otti

0.85

anton

0.84

Activations Density 0.020%

proper nouns, specifically the name "Marc"

No Comments

No Known Activations