INDEX

Explanations

terms related to geographical locations, particularly focusing on regions or directions

oai_token-act-pair · gpt-3.5-turbo

references to directional terms, specifically "north."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

•space_direction

north

No Comments

Negative Logits

ulous

-0.79

MFT

-0.74

TABLE

-0.74

aminer

-0.72

PLIC

-0.71

Information

-0.71

Own

-0.70

ellation

-0.69

roma

-0.69

 CLSID

-0.68

POSITIVE LOGITS

ward

1.36

side

1.07

wards

1.07

bound

1.04

west

1.00

ampton

0.98

shore

0.94

western

0.92

 pole

0.89

 east

0.87

Activations Density 0.036%

terms related to geographical locations, particularly focusing on regions or directions

references to directional terms, specifically "north."

No Comments

No Known Activations