INDEX

Explanations

adjectives ending in 'ional'

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of being functional or relating to institutional frameworks

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

KO

-0.79

tower

-0.78

 Primal

-0.69

uden

-0.67

âĸĪâĸĪâĸĪâĸĪ

-0.67

gone

-0.67

kok

-0.65

conom

-0.65

spell

-0.64

RGB

-0.63

POSITIVE LOGITS

ism

1.21

ized

1.16

ities

1.14

ization

1.05

ised

1.04

ists

1.01

isations

0.96

istic

0.95

idad

0.94

izational

0.94

Activations Density 0.023%

adjectives ending in 'ional'

terms related to the concept of being functional or relating to institutional frameworks

No Comments

No Known Activations