INDEX

Explanations

names of people and places

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and locations

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

âĹ¼

-0.78

lain

-0.72

ask

-0.70

realDonaldTrump

-0.69

gow

-0.66

 Boone

-0.66

selves

-0.66

ifice

-0.65

 clipboard

-0.64

 bapt

-0.63

POSITIVE LOGITS

 Fighters

1.12

ruits

0.98

ornia

0.95

letcher

0.94

ruary

0.85

avorable

0.85

rican

0.83

liction

0.83

andom

0.83

ortunate

0.80

Activations Density 0.117%

names of people and places

proper nouns, particularly names and locations

No Comments

No Known Activations