INDEX

Explanations

instances of letters addressed to specific entities or individuals

oai_token-act-pair · gpt-3.5-turbo

references to "letters to" various entities or individuals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

nels

-0.75

ahime

-0.74

 bets

-0.73

ource

-0.68

ain

-0.68

 deductions

-0.67

 smokes

-0.67

 falls

-0.65

iHUD

-0.63

wagen

-0.62

POSITIVE LOGITS

 accompany

1.13

wered

0.90

 clarify

0.87

 commemorate

0.86

pless

0.81

 signify

0.79

 address

0.79

 acquaint

0.79

 appease

0.78

ogle

0.78

Activations Density 0.121%

instances of letters addressed to specific entities or individuals

references to "letters to" various entities or individuals

No Comments

No Known Activations