INDEX

Explanations

questions with a specific formatting, likely indicating structured interviews or conversations

oai_token-act-pair · gpt-3.5-turbo

questions and inquiries involving 'Q' designations

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

 pocket

-0.75

 endanger

-0.72

olesc

-0.68

 degener

-0.67

 flock

-0.66

 continu

-0.66

 remnant

-0.63

 flood

-0.62

 purse

-0.62

 flare

-0.62

POSITIVE LOGITS

 Explain

1.30

 Speaking

1.15

How

1.14

 Were

1.14

 Congratulations

1.13

 Interesting

1.13

How

1.13

 What

1.13

Tell

1.13

 Alright

1.13

Activations Density 0.096%

questions with a specific formatting, likely indicating structured interviews or conversations

questions and inquiries involving 'Q' designations

No Comments

No Known Activations