INDEX

Explanations

phrases indicating similarity or comparison

oai_token-act-pair · gpt-3.5-turbo

comparisons expressing similarity

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

alez

-0.93

ourse

-0.87

ESA

-0.83

Ö¼

-0.83

isexual

-0.82

inion

-0.80

onding

-0.78

alt

-0.76

Cause

-0.75

ipolar

-0.74

POSITIVE LOGITS

lier

1.09

lihood

1.02

liest

0.96

 Andromeda

0.74

 crap

0.72

gib

0.70

flame

0.69

 Carth

0.66

 lifeless

0.66

liness

0.66

Activations Density 0.033%

phrases indicating similarity or comparison

comparisons expressing similarity

No Comments

No Known Activations