INDEX

Explanations

positive outcomes or situations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating well-being and positive experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

enge

-0.77

sten

-0.70

inge

-0.66

 specific

-0.64

part

-0.62

agine

-0.62

ammy

-0.60

etus

-0.60

 Clash

-0.59

iosyncr

-0.58

POSITIVE LOGITS

 except

0.99

 thanks

0.93

 despite

0.86

 smoothly

0.81

 intact

0.81

Reply

0.78

Enjoy

0.76

 alright

0.76

despite

0.76

 unchanged

0.74

Activations Density 0.351%

positive outcomes or situations

phrases indicating well-being and positive experiences

No Comments

No Known Activations