INDEX

Explanations

the word "Fram" and variants of it within the text

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Fram" or its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Downloadha

-0.72

 Elon

-0.72

REDACTED

-0.68

 Yuri

-0.64

DOE

-0.64

 Haram

-0.64

 Levine

-0.62

 Mercer

-0.62

 Azerb

-0.61

ibaba

-0.60

POSITIVE LOGITS

eworks

1.34

pton

1.17

erate

1.14

ingham

1.07

raped

0.86

eline

0.84

erity

0.84

ance

0.83

 Fram

0.83

eless

0.82

Activations Density 0.010%

the word "Fram" and variants of it within the text

occurrences of the word "Fram" or its variations

No Comments

No Known Activations