INDEX

Explanations

instances of the word "never"

oai_token-act-pair · gpt-3.5-turbo

the word "never" indicating a negative or lack of experience or occurrence

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

 Tags

-0.82

ocard

-0.75

lass

-0.74

 Inventory

-0.70

States

-0.68

uction

-0.67

GW

-0.67

endar

-0.67

achev

-0.66

ahime

-0.66

POSITIVE LOGITS

theless

1.33

 dreamed

0.90

 hesitated

0.87

 hesitate

0.85

 EVER

0.81

icably

0.81

 heard

0.79

 forgotten

0.79

 bothered

0.77

entimes

0.77

Activations Density 0.026%

instances of the word "never"

the word "never" indicating a negative or lack of experience or occurrence

No Comments

No Known Activations