INDEX

Explanations

mentions of geographical locations or place names, particularly those with "Saint" in them

oai_token-act-pair · gpt-3.5-turbo

references to places named "Saint" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

BIP

-0.89

JUST

-0.80

ptin

-0.78

PN

-0.75

PT

-0.74

iferation

-0.74

razil

-0.73

ITED

-0.70

mable

-0.68

HO

-0.68

POSITIVE LOGITS

 Laurent

0.87

 Saint

0.84

arts

0.80

 Louis

0.78

 Clair

0.78

Saint

0.77

clair

0.75

onew

0.75

 Petersburg

0.74

iless

0.73

Activations Density 0.005%

mentions of geographical locations or place names, particularly those with "Saint" in them

references to places named "Saint" in various contexts

No Comments

No Known Activations