INDEX

Explanations

words related to English-speaking countries and their cultures

oai_token-act-pair · gpt-3.5-turbo

linguistic markers related to English lexical items and their derivatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

CBC

-0.69

é¾

-0.67

 precedence

-0.65

taboola

-0.65

Bloomberg

-0.61

 applicable

-0.60

 ACTIONS

-0.60

 alloc

-0.59

nyder

-0.59

JECT

-0.59

POSITIVE LOGITS

ili

0.80

orno

0.79

atown

0.76

regor

0.76

inki

0.74

alian

0.74

adesh

0.72

ribune

0.71

urat

0.70

uay

0.70

Activations Density 0.218%

words related to English-speaking countries and their cultures

linguistic markers related to English lexical items and their derivatives

No Comments

No Known Activations