INDEX

Explanations

plural nouns

oai_token-act-pair · gpt-3.5-turbo

mentions of multiple entities or items

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

ahime

-0.66

 Activities

-0.64

SPONSORED

-0.63

 Leth

-0.63

Fre

-0.62

rum

-0.62

 Anthrop

-0.62

 Boxing

-0.61

 Newsletter

-0.60

humane

-0.60

POSITIVE LOGITS

 apiece

1.21

 totaling

1.01

poons

0.93

paces

0.90

hips

0.84

 thirds

0.79

 simultaneously

0.77

ixt

0.75

 consecut

0.75

agos

0.74

Activations Density 0.255%

plural nouns

mentions of multiple entities or items

No Comments

No Known Activations