INDEX

Explanations

adjectives ending in '-ness' to describe a characteristic or quality

oai_token-act-pair · gpt-3.5-turbo

terms related to various states or qualities characterized by the suffix 'ness'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

odium

-0.75

herty

-0.75

vice

-0.72

bsite

-0.70

anski

-0.66

inav

-0.66

roleum

-0.66

nesium

-0.65

umm

-0.65

ãĥ¯ãĥ³

-0.63

POSITIVE LOGITS

iness

0.97

 quot

0.90

 Principle

0.87

uality

0.79

lessness

0.78

phal

0.78

fulness

0.76

yy

0.76

 bordering

0.75

 nerv

0.74

Activations Density 0.145%

adjectives ending in '-ness' to describe a characteristic or quality

terms related to various states or qualities characterized by the suffix 'ness'

No Comments

No Known Activations