INDEX

Explanations

proper nouns or specific terms related to organizations or locations

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and specifiers in context

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

 guiActiveUn

-0.81

ngth

-0.80

ĸļ

-0.71

igraph

-0.70

ometimes

-0.64

osterone

-0.63

OPLE

-0.62

stanbul

-0.62

ugu

-0.61

æĥ

-0.61

POSITIVE LOGITS

lehem

0.99

pillar

0.89

levard

0.85

bodied

0.80

mingham

0.76

abase

0.68

apest

0.67

oks

0.67

ĵĺ

0.67

dylib

0.65

Activations Density 0.050%

proper nouns or specific terms related to organizations or locations

punctuation marks and specifiers in context

No Comments

No Known Activations