INDEX

Explanations

terms related to government, authority, and power

oai_token-act-pair · gpt-3.5-turbo

suffixes related to adjectives and nouns

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

 IMAGES

-0.82

merce

-0.81

 Slate

-0.75

ãģ®éŃĶ

-0.65

``

-0.64

arrell

-0.61

ÏĢ

-0.60

áµ

-0.60

BALL

-0.59

LET

-0.59

POSITIVE LOGITS

ndum

0.85

ardless

0.84

ativity

0.73

eering

0.70

omez

0.69

owship

0.69

ency

0.69

idated

0.69

andom

0.68

alg

0.68

Activations Density 0.085%

terms related to government, authority, and power

suffixes related to adjectives and nouns

No Comments

No Known Activations