INDEX

Explanations

positive evaluations or comments

oai_token-act-pair · gpt-3.5-turbo

phrases expressing affirmation or positive sentiment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

pora

-0.75

gypt

-0.74

cro

-0.73

 guiActiveUn

-0.72

ural

-0.68

ascript

-0.66

acan

-0.66

arij

-0.65

atari

-0.64

iless

-0.64

POSITIVE LOGITS

 albeit

0.67

âĹ¼

0.67

 congratulations

0.66

 Reson

0.62

Brave

0.59

 outweigh

0.58

 fodder

0.58

 insofar

0.57

albeit

0.57

hearted

0.57

Activations Density 0.561%

positive evaluations or comments

phrases expressing affirmation or positive sentiment

No Comments

No Known Activations