INDEX

Explanations

names related to people or places

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Bar" followed by various identifiers or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

lihood

-0.93

SOS

-0.74

ãģį

-0.70

UAL

-0.70

hower

-0.69

IBLE

-0.68

 Dangerous

-0.67

 STATES

-0.66

å£«

-0.66

CRIP

-0.65

POSITIVE LOGITS

becue

1.23

riers

1.16

celona

1.16

itone

1.16

bell

1.06

rington

1.04

rier

1.04

bara

1.00

iatric

0.98

keep

0.96

Activations Density 0.017%

names related to people or places

occurrences of the name "Bar" followed by various identifiers or titles

No Comments

No Known Activations