INDEX

Explanations

keywords related to news and updates

oai_token-act-pair · gpt-3.5-turbo

mentions of news or media-related content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Osw

-0.61

 Logged

-0.60

 misunder

-0.59

ovember

-0.59

lished

-0.58

igi

-0.58

ukong

-0.57

asus

-0.57

zbollah

-0.56

anza

-0.55

POSITIVE LOGITS

worthiness

0.79

worthy

0.76

letters

0.74

flash

0.70

packages

0.67

feed

0.67

 literacy

0.67

 content

0.66

alg

0.65

 Kavanaugh

0.65

Activations Density 0.012%

keywords related to news and updates

mentions of news or media-related content

No Comments

No Known Activations