INDEX

Explanations

locations or landmarks related to bodies of water

oai_token-act-pair · gpt-3.5-turbo

mentions of specific harbors

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

uable

-0.85

ivia

-0.84

acca

-0.80

umar

-0.76

chers

-0.73

Downloadha

-0.73

oln

-0.72

DonaldTrump

-0.72

okers

-0.71

ongyang

-0.71

POSITIVE LOGITS

 Harbor

0.92

 Ship

0.81

 Island

0.81

 Plaza

0.78

 Shore

0.78

 Harbour

0.76

 Islands

0.74

 Yard

0.73

front

0.73

 Islanders

0.72

Activations Density 0.007%

locations or landmarks related to bodies of water

mentions of specific harbors

No Comments

No Known Activations