INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

instances of the word "said."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

clad

-0.67

 advoc

-0.67

 conflic

-0.63

intern

-0.63

oppable

-0.61

æĸ¹

-0.61

 arrang

-0.60

apult

-0.58

emed

-0.57

 shenan

-0.57

POSITIVE LOGITS

 hello

0.71

 msec

0.71

 goodbye

0.71

=\"

0.61

Psy

0.59

 href

0.57

rius

0.57

0.54

ys

0.54

 farewell

0.53

Activations Density 0.050%

proper nouns

instances of the word "said."

No Comments

No Known Activations