INDEX

Explanations

words related to starting, initializing, or beginning actions or processes

oai_token-act-pair · gpt-3.5-turbo

terms related to starting processes or actions

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

aver

-0.75

Chance

-0.68

 Truck

-0.67

umbo

-0.66

 Gould

-0.66

Die

-0.62

 Extrem

-0.61

avering

-0.60

 Machina

-0.60

Hom

-0.60

POSITIVE LOGITS

 initi

1.16

 initiated

1.06

 initiate

0.96

 initiation

0.93

TAIN

0.88

 initiating

0.86

chnology

0.83

phase

0.82

rative

0.79

wcs

0.76

Activations Density 0.016%

words related to starting, initializing, or beginning actions or processes

terms related to starting processes or actions

No Comments

No Known Activations