INDEX

Explanations

adjectives related to negative qualities or conditions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "weak" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ittee

-0.75

 Ready

-0.71

 Newly

-0.68

 McKenna

-0.68

CENT

-0.67

ICAN

-0.67

 Cycling

-0.67

OGR

-0.67

 Hilton

-0.67

rection

-0.65

POSITIVE LOGITS

nesses

1.25

 weak

1.07

 weakest

0.99

ens

0.96

luster

0.93

lings

0.92

 weakness

0.86

 weaker

0.86

ening

0.86

 undermin

0.85

Activations Density 0.006%

adjectives related to negative qualities or conditions

instances of the word "weak" in various contexts

No Comments

No Known Activations