INDEX

Explanations

the word "attempt" or its variations with high activation

oai_token-act-pair · gpt-3.5-turbo

phrases related to attempts or efforts made in various contexts

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

eyes

-0.61

head

-0.59

houses

-0.58

 Wrap

-0.58

 Angola

-0.58

baum

-0.57

effect

-0.56

wine

-0.55

ranging

-0.55

 Blend

-0.55

POSITIVE LOGITS

Attempts

1.01

 unsuccessfully

0.95

to

0.86

llor

0.77

 attempts

0.74

 attempt

0.72

bles

0.71

at

0.68

 succeed

0.68

 futile

0.67

Activations Density 0.051%

the word "attempt" or its variations with high activation

phrases related to attempts or efforts made in various contexts

No Comments

No Known Activations