INDEX

Explanations

unique character names and settings within a story

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and their associated context

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

abwe

-0.78

 welding

-0.76

 withdraw

-0.75

 medicines

-0.72

 lifes

-0.72

 inflation

-0.71

 furnace

-0.71

wagen

-0.71

 weld

-0.68

 withdrawal

-0.68

POSITIVE LOGITS

Advertisement

1.09

Speaking

1.00

Anyway

0.99

pmwiki

0.98

Favorite

0.96

Disney

0.94

Film

0.90

Screen

0.90

PHOTOS

0.88

Elsewhere

0.88

Activations Density 0.556%

unique character names and settings within a story

punctuation marks and their associated context

No Comments

No Known Activations