INDEX

Explanations

the word "mind."

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "mind" or mental processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Mehran

-0.73

Recomm

-0.65

 Aval

-0.64

Pyr

-0.63

ICAN

-0.62

INO

-0.61

 Canter

-0.61

 Bunker

-0.60

TAIN

-0.59

 Restaur

-0.59

POSITIVE LOGITS

fulness

1.26

share

1.19

storms

1.17

fuck

1.10

bender

1.05

scape

1.05

lessly

1.04

ets

1.03

sets

0.97

trak

0.97

Activations Density 0.043%

the word "mind."

references to the concept of "mind" or mental processes

No Comments

No Known Activations