INDEX

Explanations

numbers with the suffix 'plus'

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a duration of "plus" years or decades

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

inem

-0.77

arij

-0.69

ayn

-0.68

omach

-0.67

adr

-0.66

ringe

-0.66

rouch

-0.66

icht

-0.65

bane

-0.64

oscope

-0.63

POSITIVE LOGITS

plus

0.98

minus

0.95

--+

0.84

Plus

0.83

ishly

0.80

/-

0.79

------------

0.74

cules

0.73

 Scotia

0.73

ãģĨ

0.71

Activations Density 0.006%

numbers with the suffix 'plus'

phrases indicating a duration of "plus" years or decades

No Comments

No Known Activations