INDEX

Explanations

mentions of geographical locations or areas

oai_token-act-pair · gpt-3.5-turbo

references to specific land entities or regions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ople

-0.71

 surgical

-0.65

vernment

-0.65

ilies

-0.64

ATH

-0.64

ascus

-0.64

Reason

-0.63

 dism

-0.61

Visual

-0.61

 numerical

-0.60

POSITIVE LOGITS

bucks

1.07

meet

0.95

 lands

0.90

hips

0.90

dale

0.88

holder

0.88

marks

0.88

kamp

0.84

cence

0.81

lands

0.77

Activations Density 0.011%

mentions of geographical locations or areas

references to specific land entities or regions

No Comments

No Known Activations