INDEX

Explanations

words related to specific place names

oai_token-act-pair · gpt-3.5-turbo

proper nouns, especially names of people and places

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

enhagen

-0.71

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.68

 entrance

-0.62

 priceless

-0.61

 vide

-0.60

ratom

-0.59

draft

-0.59

 imposing

-0.59

 admission

-0.58

 blaster

-0.58

POSITIVE LOGITS

anooga

0.89

nai

0.86

leon

0.85

aign

0.82

obos

0.82

hei

0.81

ulhu

0.79

esy

0.79

lain

0.78

 Disciple

0.74

Activations Density 0.131%

words related to specific place names

proper nouns, especially names of people and places

No Comments

No Known Activations