INDEX

Explanations

proper nouns representing people

oai_token-act-pair · gpt-3.5-turbo

references to individuals or pronouns indicating people

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

 feature

-0.71

 features

-0.67

 stem

-0.63

 packed

-0.62

 preparation

-0.61

enc

-0.61

 primitive

-0.61

 ends

-0.60

 depress

-0.60

 storage

-0.59

POSITIVE LOGITS

who

3.74

whose

2.47

Who

1.81

WHO

1.79

 whom

1.78

who

1.72

how

1.49

where

1.43

Who

1.41

WHO

1.38

Activations Density 0.017%

proper nouns representing people

references to individuals or pronouns indicating people

No Comments

No Known Activations