INDEX

Explanations

prepositions denoting a location or direction

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "in"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Cursed

-0.76

 Cock

-0.69

hack

-0.64

 Charlottesville

-0.62

 Seah

-0.61

 Debor

-0.61

 Frie

-0.61

 Parenthood

-0.61

 Bangl

-0.60

 Ethnic

-0.60

POSITIVE LOGITS

iors

0.75

ifice

0.71

aples

0.70

nery

0.69

vier

0.69

viron

0.67

 BELOW

0.66

uin

0.65

cific

0.64

atos

0.64

Activations Density 0.000%

prepositions denoting a location or direction

occurrences of the word "in"

No Comments

No Known Activations