INDEX

Explanations

words related to elusive or escaping things

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "elude."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

nant

-0.71

itives

-0.68

stress

-0.67

½

-0.65

LESS

-0.64

pled

-0.61

 shatter

-0.61

xit

-0.60

llers

-0.59

ELS

-0.59

POSITIVE LOGITS

izabeth

1.24

ixir

1.14

ibrary

1.10

iquid

1.04

usive

1.03

abor

0.92

uded

0.89

ijah

0.89

bows

0.87

igible

0.87

Activations Density 0.023%

words related to elusive or escaping things

mentions of the word "elude."

No Comments

No Known Activations