INDEX

Explanations

phrases instructing to pay attention to specific things or events

oai_token-act-pair · gpt-3.5-turbo

phrases related to watching or monitoring something attentively

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

venge

-0.80

 Gibbs

-0.66

 Guitar

-0.65

HAHAHAHA

-0.63

Grade

-0.62

bet

-0.61

 Whites

-0.61

DNA

-0.61

hens

-0.61

ĪĴ

-0.60

POSITIVE LOGITS

âĦ¢:

0.74

 ahead

0.73

eful

0.73

gency

0.71

ulatory

0.70

emark

0.70

ãĤ¶

0.67

oin

0.67

agascar

0.66

uum

0.66

Activations Density 0.031%

phrases instructing to pay attention to specific things or events

phrases related to watching or monitoring something attentively

No Comments

No Known Activations