INDEX

Explanations

names of locations or places

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly city names and notable entities

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

 guiActiveUnfocused

-0.78

[&

-0.78

 thereof

-0.69

 thro

-0.69

Cth

-0.63

 Cthulhu

-0.62

 thereto

-0.60

isSpecialOrderable

-0.59

Teg

-0.59

 Osiris

-0.59

POSITIVE LOGITS

anyahu

0.99

elligence

0.96

asury

0.95

ropolitan

0.94

mond

0.93

odore

0.93

jamin

0.92

rique

0.91

ctions

0.90

alyst

0.90

Activations Density 0.220%

names of locations or places

proper nouns, particularly city names and notable entities

No Comments

No Known Activations