INDEX

Explanations

words related to communication and dialogue

oai_token-act-pair · gpt-3.5-turbo

phrases related to dialogue or speech

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 dudes

-0.66

 shitty

-0.65

 shit

-0.62

 wanna

-0.62

 crap

-0.62

 swat

-0.62

 bashing

-0.62

 trolling

-0.62

 dude

-0.61

 titan

-0.58

POSITIVE LOGITS

ogether

1.00

sequently

0.95

Therefore

0.94

 Finally

0.83

Furthermore

0.83

 Lastly

0.82

Finally

0.82

 furthermore

0.78

Moreover

0.78

 Afterwards

0.77

Activations Density 1.082%

words related to communication and dialogue

phrases related to dialogue or speech

No Comments

No Known Activations