INDEX

Explanations

medical terminology and emergency situations

oai_token-act-pair · gpt-3.5-turbo

words related to risk and vulnerability

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.56

udeb

-0.53

peat

-0.52

present

-0.51

window

-0.50

etta

-0.48

idden

-0.48

ensis

-0.48

dfx

-0.47

 impart

-0.47

POSITIVE LOGITS

 Lines

0.74

 Functions

0.73

 Directory

0.70

 Strategies

0.70

 Benefits

0.70

 Values

0.69

 Differences

0.69

 Recomm

0.68

 Advice

0.67

 Explan

0.66

Activations Density 0.757%

medical terminology and emergency situations

words related to risk and vulnerability

No Comments

No Known Activations