INDEX

Explanations

mentions of technology companies and their products

oai_token-act-pair · gpt-3.5-turbo

disconnected phrases or punctuation that suggest a shift in context or topics

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 skelet

-0.82

 neighb

-0.82

 prosec

-0.81

 quir

-0.80

 adolescence

-0.78

 enriched

-0.75

 covenant

-0.75

 crochet

-0.75

 cler

-0.74

 perfume

-0.74

POSITIVE LOGITS

However

1.94

According

1.88

Unfortunately

1.80

Therefore

1.79

While

1.73

Furthermore

1.72

Firstly

1.72

Interestingly

1.72

Below

1.71

Currently

1.69

Activations Density 0.452%

mentions of technology companies and their products

disconnected phrases or punctuation that suggest a shift in context or topics

No Comments

No Known Activations