INDEX

Explanations

words related to leadership positions or titles

oai_token-act-pair · gpt-3.5-turbo

occurrences of words or phrases that convey the concept of entrenchment or deep involvement

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

 Pyramid

-0.70

 Dusk

-0.68

 Pulse

-0.66

 Kills

-0.64

 Tsukuyomi

-0.64

 Catalyst

-0.63

 Matter

-0.62

 Heat

-0.62

resso

-0.62

 Kitchen

-0.60

POSITIVE LOGITS

TABLE

0.72

Ŀ

0.71

âĸ¬

0.71

operation

0.70

ously

0.68

sic

0.68

¿

0.68

âķ

0.68

Ãĥ

0.67

orial

0.67

Activations Density 0.145%

words related to leadership positions or titles

occurrences of words or phrases that convey the concept of entrenchment or deep involvement

No Comments

No Known Activations