INDEX

Explanations

locations or regions, specifically focusing on directional indicators (such as northern or southern)

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Sharp

-0.84

ynt

-0.83

enable

-0.79

Money

-0.77

roo

-0.77

Room

-0.76

Frames

-0.75

Filter

-0.74

neutral

-0.74

Boo

-0.73

POSITIVE LOGITS

most

1.55

 hemisphere

1.36

 regions

1.09

 suburbs

1.06

 provinces

1.01

 region

1.01

 extrem

0.97

 portion

0.92

 edge

0.92

 periphery

0.90

Activations Density 0.046%

locations or regions, specifically focusing on directional indicators (such as northern or southern)

No Comments

No Known Activations