INDEX

Explanations

keywords related to news or information roundups

oai_token-act-pair · gpt-3.5-turbo

references to news roundups or summaries

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

aternal

-0.65

cession

-0.65

hood

-0.64

ippi

-0.62

akin

-0.61

wagen

-0.61

entity

-0.61

oji

-0.60

Ã±a

-0.60

liter

-0.60

POSITIVE LOGITS

 roundup

1.42

 Roundup

0.88

weekly

0.77

endium

0.73

ãĥ¼ãĥ³

0.72

ÃįÃį

0.71

ãĥ¢

0.71

ãĤ¡

0.69

 GOODMAN

0.67

 Meal

0.67

Activations Density 0.008%

keywords related to news or information roundups

references to news roundups or summaries

No Comments

No Known Activations