INDEX

Explanations

mentions of the publication "Jyllands-Posten" and related topics

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names associated with articles or opinion pieces

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

perature

-0.83

raviolet

-0.71

asking

-0.70

rers

-0.70

legates

-0.67

holding

-0.67

packing

-0.67

positive

-0.66

fixed

-0.65

SHIP

-0.64

POSITIVE LOGITS

shire

0.99

yll

0.94

sburgh

0.78

sburg

0.77

yd

0.76

odge

0.74

anes

0.73

INGTON

0.70

oyd

0.70

ic

0.70

Activations Density 0.011%

mentions of the publication "Jyllands-Posten" and related topics

proper nouns, particularly names associated with articles or opinion pieces

No Comments

No Known Activations