INDEX

Explanations

references to the word "liter" or variations of it

oai_token-act-pair · gpt-3.5-turbo

terms related to literacy

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

cffffcc

-0.77

 Sands

-0.69

 Dominion

-0.68

 Secret

-0.64

EStream

-0.64

 Fate

-0.64

Agent

-0.63

 visit

-0.62

Faw

-0.62

querade

-0.59

POSITIVE LOGITS

 liter

1.40

liter

1.38

atures

1.22

 Liter

1.15

Liter

0.98

acies

0.98

 illiter

0.97

umen

0.96

uania

0.94

ature

0.87

Activations Density 0.009%

references to the word "liter" or variations of it

terms related to literacy

No Comments

No Known Activations