INDEX

Explanations

personal and location names, especially "Shore" and "Weston."

oai_token-act-pair · gpt-3.5-turbo

references to specific locations or names associated with places

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

ivia

-0.94

ibles

-0.92

IBLE

-0.92

ible

-0.91

orescent

-0.85

ibility

-0.84

elta

-0.84

untu

-0.84

orate

-0.79

umbnail

-0.78

POSITIVE LOGITS

 Shore

0.97

stals

0.90

bone

0.89

pool

0.83

line

0.83

hoff

0.83

castle

0.82

craft

0.81

brook

0.81

ham

0.80

Activations Density 0.022%

personal and location names, especially "Shore" and "Weston."

references to specific locations or names associated with places

No Comments

No Known Activations