INDEX

Explanations

names and titles, potentially related to important or influential figures

oai_token-act-pair · gpt-3.5-turbo

names of significant individuals or titles related to leadership and authority

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

ibles

-0.93

ĪĴ

-0.89

etics

-0.84

etheus

-0.78

cible

-0.78

lihood

-0.78

etic

-0.77

pedia

-0.77

etically

-0.75

esis

-0.73

POSITIVE LOGITS

sth

0.82

bara

0.80

 Mahar

0.74

Ki

0.69

dos

0.68

dinand

0.67

Govern

0.67

KB

0.64

WER

0.64

Pai

0.64

Activations Density 0.024%

names and titles, potentially related to important or influential figures

names of significant individuals or titles related to leadership and authority

No Comments

No Known Activations