INDEX

Explanations

phrases or terms related to the concept of starting or initiating something

oai_token-act-pair · gpt-3.5-turbo

references to "start" or "startup" concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

illard

-0.87

otropic

-0.72

ochem

-0.71

phy

-0.71

ournal

-0.70

ugs

-0.69

 delicate

-0.68

ocally

-0.68

 autopsy

-0.67

idal

-0.67

POSITIVE LOGITS

starting

1.06

start

1.03

watch

0.97

ners

0.97

Start

0.96

 Start

0.96

strap

0.85

nings

0.82

lace

0.80

Stars

0.80

Activations Density 0.016%

phrases or terms related to the concept of starting or initiating something

references to "start" or "startup" concepts

No Comments

No Known Activations