INDEX

Explanations

proper nouns, specifically the acronym "DW."

oai_token-act-pair · gpt-3.5-turbo

references to the German broadcaster DW and related content

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

creen

-0.75

Magikarp

-0.74

ozo

-0.72

olithic

-0.67

eers

-0.67

ãĥ¼ãĥĨ

-0.67

ramid

-0.65

inates

-0.64

eering

-0.64

ocene

-0.64

POSITIVE LOGITS

DW

1.09

DW

1.08

ORD

1.02

ND

0.87

ried

0.86

TD

0.83

BY

0.82

IGHT

0.81

LP

0.78

LM

0.77

Activations Density 0.008%

proper nouns, specifically the acronym "DW."

references to the German broadcaster DW and related content

No Comments

No Known Activations