INDEX

Explanations

designated entities or areas

oai_token-act-pair · gpt-3.5-turbo

instances of the word "designated"!

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

stead

-0.68

conom

-0.66

corn

-0.64

Fish

-0.62

riger

-0.62

athi

-0.61

Rh

-0.60

Fal

-0.59

ENG

-0.59

gain

-0.59

POSITIVE LOGITS

 designated

0.89

etheless

0.75

 designate

0.74

 warr

0.70

 designation

0.69

 markings

0.69

İĭ

0.68

 Parenthood

0.67

ependent

0.65

orsche

0.64

Activations Density 0.027%

designated entities or areas

instances of the word "designated"!

No Comments

No Known Activations