INDEX

Explanations

words related to strokes, specifically "stro" in various forms such as "stroke" and "stroking."

oai_token-act-pair · gpt-3.5-turbo

variations of the word "stroke."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 reviewer

-0.64

ãģ®éŃĶ

-0.62

æĺ¯

-0.61

arian

-0.60

spring

-0.60

ARDIS

-0.60

arians

-0.60

illary

-0.60

NPR

-0.59

 Forbidden

-0.59

POSITIVE LOGITS

kes

1.07

stro

1.05

kered

0.92

zzy

0.92

otom

0.91

oshenko

0.91

vous

0.90

bered

0.89

bles

0.88

usalem

0.86

Activations Density 0.012%

words related to strokes, specifically "stro" in various forms such as "stroke" and "stroking."

variations of the word "stroke."

No Comments

No Known Activations