INDEX

Explanations

numbers that are part of measurements or quantities

oai_token-act-pair · gpt-3.5-turbo

numerical data or statistics

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

bara

-0.81

wart

-0.73

money

-0.73

egal

-0.70

skirts

-0.68

arya

-0.67

oras

-0.67

adder

-0.67

estine

-0.66

actic

-0.66

POSITIVE LOGITS

lbs

0.86

 pounds

0.84

 ILCS

0.81

lbs

0.78

MHz

0.76

MHz

0.76

 calories

0.75

OPLE

0.73

FY

0.73

 degrees

0.70

Activations Density 0.084%

numbers that are part of measurements or quantities

numerical data or statistics

No Comments

No Known Activations