INDEX

Explanations

words related to excellence in communication or strong beliefs and feelings

oai_token-act-pair · gpt-3.5-turbo

words or phrases that convey strong emotional commitment or eloquence

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

 Redemption

-0.69

 spotting

-0.69

 Dahl

-0.69

 Oval

-0.68

hyde

-0.66

 Sund

-0.65

 hatch

-0.64

LCS

-0.63

bay

-0.63

places

-0.62

POSITIVE LOGITS

uously

1.57

ently

1.36

aciously

1.31

orously

1.24

antly

1.21

uous

1.20

cially

1.16

uably

1.15

iously

1.15

ously

1.14

Activations Density 0.065%

words related to excellence in communication or strong beliefs and feelings

words or phrases that convey strong emotional commitment or eloquence

No Comments

No Known Activations