INDEX

Explanations

years

oai_token-act-pair · gpt-3.5-turbo

references to the year 1973

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ndra

-0.77

tro

-0.75

hop

-0.71

Edge

-0.69

venge

-0.69

BOOK

-0.68

erate

-0.67

Storage

-0.66

 entirety

-0.66

ventus

-0.65

POSITIVE LOGITS

 onwards

0.74

 onward

0.69

ĸļ

0.68

0.66

0.65

å¹

0.64

 Jong

0.63

 Chevrolet

0.63

 ILCS

0.63

Activations Density 0.032%

years

references to the year 1973

No Comments

No Known Activations