INDEX

Explanations

words related to locations or entities with a common "lo" sequence

oai_token-act-pair · gpt-3.5-turbo

words that end with specific suffixes

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĪĴ

-0.91

ADRA

-0.69

 looph

-0.69

displayText

-0.67

catentry

-0.67

ERO

-0.66

dayName

-0.66

CRIP

-0.65

llah

-0.65

 tremend

-0.65

POSITIVE LOGITS

rences

0.88

glers

0.81

warm

0.79

iberal

0.77

nder

0.73

hood

0.73

andestine

0.72

omorphic

0.71

ocket

0.70

Ã©s

0.70

Activations Density 0.039%

words related to locations or entities with a common "lo" sequence

words that end with specific suffixes

No Comments

No Known Activations