INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of individuals or entities

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

acebook

-0.71

Â·Â·

-0.71

LEASE

-0.68

ãĤ¨ãĥ«

-0.64

PsyNetMessage

-0.64

Animal

-0.64

âĶĢâĶĢ

-0.63

 Cheryl

-0.60

IUM

-0.58

LCS

-0.58

POSITIVE LOGITS

utsch

1.01

oub

0.84

deen

0.84

iman

0.84

zynski

0.83

iso

0.82

uy

0.81

kil

0.81

iani

0.80

chu

0.80

Activations Density 0.299%

names of individuals

proper nouns, particularly names of individuals or entities

No Comments

No Known Activations