INDEX

Explanations

words related to items or objects that are widely used or consumed

oai_token-act-pair · gpt-3.5-turbo

nouns that refer to various types of professionals or roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ADRA

-0.71

URRENT

-0.69

ĸļ

-0.68

 Foss

-0.67

 forfeiture

-0.65

 numerical

-0.65

ATIONAL

-0.64

orsche

-0.60

 conviction

-0.59

 Downing

-0.58

POSITIVE LOGITS

mith

1.17

haw

1.10

paces

1.06

ucker

1.01

leeve

0.99

cream

0.96

hots

0.96

ystem

0.95

cale

0.94

pace

0.94

Activations Density 0.068%

words related to items or objects that are widely used or consumed

nouns that refer to various types of professionals or roles

No Comments

No Known Activations