INDEX

Explanations

keywords referring to specific objects or entities within a broader context

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or items denoted as "ones."

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

Cl

-0.69

Phys

-0.67

Dep

-0.64

inson

-0.63

osponsors

-0.62

Ec

-0.61

amen

-0.61

ITED

-0.61

Pwr

-0.61

Dem

-0.60

POSITIVE LOGITS

hots

0.87

omething

0.80

 eyed

0.74

cott

0.73

 Hundred

0.72

 Thousand

0.70

elf

0.69

ones

0.69

chwitz

0.68

Obi

0.67

Activations Density 0.034%

keywords referring to specific objects or entities within a broader context

references to specific entities or items denoted as "ones."

No Comments

No Known Activations