INDEX

Explanations

phrases related to spatial locations or movements

oai_token-act-pair · gpt-3.5-turbo

references to geographical locations, specifically focusing on "the" followed by a location name

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

uala

-0.76

 Canaver

-0.73

ourney

-0.72

%%%%

-0.70

ardon

-0.67

azel

-0.66

 anyway

-0.66

umm

-0.65

NB

-0.64

 nonetheless

-0.64

POSITIVE LOGITS

 sexes

1.35

 extremes

0.83

 poles

0.81

 respective

0.81

 aforementioned

0.79

 factions

0.79

 genders

0.78

 combatants

0.78

 parties

0.77

 dots

0.73

Activations Density 0.117%

phrases related to spatial locations or movements

references to geographical locations, specifically focusing on "the" followed by a location name

No Comments

No Known Activations