INDEX

Explanations

phrases related to hierarchical structures or organizations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating rank-and-file dynamics or relationships within organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Newsletter

-0.71

HT

-0.65

 Helpful

-0.60

":""},{"

-0.59

hn

-0.59

enegger

-0.59

Wallet

-0.59

ãĤ¯

-0.58

Reloaded

-0.58

nw

-0.57

POSITIVE LOGITS

 parcel

0.84

 likeness

0.77

 circumstance

0.71

 match

0.71

 gown

0.71

 conquer

0.68

lege

0.66

 Conquer

0.64

 fury

0.63

practice

0.62

Activations Density 0.135%

phrases related to hierarchical structures or organizations

phrases indicating rank-and-file dynamics or relationships within organizations

No Comments

No Known Activations