INDEX

Explanations

locations or settings within a text

oai_token-act-pair · gpt-3.5-turbo

situations involving physical entrapment or confinement

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

 Seller

-0.61

rosso

-0.59

Reviewer

-0.58

Pwr

-0.56

Commerce

-0.55

eller

-0.55

DragonMagazine

-0.54

=-=-=-=-

-0.54

Trade

-0.54

ennial

-0.53

POSITIVE LOGITS

 startled

0.72

 throats

0.69

stretched

0.69

 unconscious

0.67

 barrier

0.67

 shoulder

0.65

 choke

0.64

 windshield

0.64

 doorway

0.63

 shaft

0.62

Activations Density 0.902%

locations or settings within a text

situations involving physical entrapment or confinement

No Comments

No Known Activations