INDEX

Explanations

references to negative events or conditions

oai_token-act-pair · gpt-3.5-turbo

negative news or reports about unfavorable situations

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

sonian

-0.92

çīĪ

-0.82

 dexter

-0.76

£ı

-0.75

 confidently

-0.75

verning

-0.72

cyclop

-0.70

 trustworthy

-0.70

encers

-0.70

ritz

-0.70

POSITIVE LOGITS

 Karma

0.79

:(

0.70

 Syndrome

0.70

 Clown

0.68

ãĤ®

0.67

 Worse

0.67

 miser

0.66

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

0.66

 incurred

0.66

 imaginable

0.65

Activations Density 0.373%

references to negative events or conditions

negative news or reports about unfavorable situations

No Comments

No Known Activations