INDEX

Explanations

positive or affirmative words

oai_token-act-pair · gpt-3.5-turbo

expressions of positivity or affirmations related to well-being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

apsed

-0.74

igan

-0.74

aciously

-0.72

 Serpent

-0.68

 coma

-0.68

hurst

-0.66

xual

-0.65

isexual

-0.64

ition

-0.64

lust

-0.63

POSITIVE LOGITS

Good

2.20

Excellent

1.01

ORPG

0.93

uo

0.90

Broad

0.89

 Goodbye

0.87

Wo

0.87

 livest

0.82

Whe

0.79

uity

0.78

Activations Density 0.026%

positive or affirmative words

expressions of positivity or affirmations related to well-being

No Comments

No Known Activations