INDEX

Explanations

questions about actions or situations

oai_token-act-pair · gpt-3.5-turbo

inquiries or questions about actions or situations involving "what are you doing."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 synerg

-0.69

arcity

-0.68

 significantly

-0.65

 councils

-0.61

 notably

-0.60

 markedly

-0.60

 mentors

-0.60

 nodes

-0.59

 attest

-0.57

 broadly

-0.57

POSITIVE LOGITS

?!"

1.12

?!

1.06

!?

0.98

 gonna

0.98

!?"

0.97

?'

0.90

...?

0.89

?"

0.87

?'"

0.87

?",

0.86

Activations Density 0.172%

questions about actions or situations

inquiries or questions about actions or situations involving "what are you doing."

No Comments

No Known Activations