INDEX

Explanations

references to 'Me' or activities involving the identification or measurement of entities

oai_token-act-pair · gpt-3.5-turbo

mentions of the term "Me," indicating a focus on personal narrative or experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

raper

-0.74

itialized

-0.69

trump

-0.66

medium

-0.65

çĶŁ

-0.64

*/(

-0.62

raviolet

-0.62

 vacuum

-0.62

paralle

-0.61

 Jindal

-0.61

POSITIVE LOGITS

eting

1.04

asures

1.03

zzo

1.03

asure

1.00

asured

0.96

cca

0.95

asuring

0.93

aning

0.93

adows

0.89

adow

0.83

Activations Density 0.013%

references to 'Me' or activities involving the identification or measurement of entities

mentions of the term "Me," indicating a focus on personal narrative or experiences

No Comments

No Known Activations