INDEX

Explanations

words related to positive impact or benefit

oai_token-act-pair · gpt-3.5-turbo

the concept of "good" in various contexts

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

eters

-0.79

agos

-0.74

ptin

-0.73

hod

-0.70

âĹ¼

-0.67

 Pavilion

-0.67

pper

-0.67

kson

-0.66

ocene

-0.66

gow

-0.65

POSITIVE LOGITS

enough

1.10

reads

1.05

 deed

0.95

 intentions

0.94

 deeds

0.93

 Samar

0.92

 luck

0.91

sword

0.89

NESS

0.81

luck

0.81

Activations Density 0.056%

words related to positive impact or benefit

the concept of "good" in various contexts

No Comments

No Known Activations