INDEX

Explanations

words related to definitions or precise descriptions

oai_token-act-pair · gpt-3.5-turbo

the prefix "def" which typically indicates definitions or terms being defined

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

externalActionCode

-0.80

DAY

-0.71

sth

-0.69

[|

-0.68

STON

-0.67

 Boll

-0.67

 Dragonbound

-0.66

assies

-0.66

HAHA

-0.64

Pom

-0.64

POSITIVE LOGITS

initions

1.49

initely

1.43

icient

1.17

ence

1.16

inately

1.09

ensible

1.08

inite

1.08

ensive

1.06

eat

1.04

endant

1.03

Activations Density 0.009%

words related to definitions or precise descriptions

the prefix "def" which typically indicates definitions or terms being defined

No Comments

No Known Activations