INDEX

Explanations

proper nouns or specific entities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Found" with varying levels of significance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 spoiler

-0.78

 schedule

-0.75

 override

-0.70

 electronically

-0.67

 reel

-0.67

 ridicule

-0.66

 divert

-0.66

 withholding

-0.66

 tense

-0.65

 manual

-0.65

POSITIVE LOGITS

 Found

3.71

Found

3.34

found

2.24

 Reached

1.52

 foundations

1.41

 Identified

1.34

 Noticed

1.34

 Fell

1.30

 found

1.26

 Bought

1.25

Activations Density 0.012%

proper nouns or specific entities

occurrences of the word "Found" with varying levels of significance

No Comments

No Known Activations