INDEX

Explanations

mentions of "ters" or words ending in "ters."

oai_token-act-pair · gpt-3.5-turbo

words related to various forms of "ter" and "tering."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Shar

-0.67

IRO

-0.59

Graphics

-0.59

Ĥİ

-0.59

Bah

-0.58

ILLE

-0.58

Fra

-0.57

 reconstruction

-0.57

 circum

-0.56

Â¢

-0.55

POSITIVE LOGITS

weet

1.00

geist

0.91

ranean

0.85

rible

0.83

ified

0.82

rors

0.81

rane

0.80

tainment

0.80

theless

0.80

lie

0.77

Activations Density 0.015%

mentions of "ters" or words ending in "ters."

words related to various forms of "ter" and "tering."

No Comments

No Known Activations