INDEX

Explanations

pronouns and verbs indicating self-directed actions

oai_token-act-pair · gpt-3.5-turbo

references to personal pronouns and the concept of self

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

ulton

-0.68

 Clover

-0.65

Mub

-0.65

vertisements

-0.63

 Availability

-0.63

 Alger

-0.62

iens

-0.62

ritz

-0.62

heny

-0.61

microsoft

-0.61

POSITIVE LOGITS

selves

0.96

 selves

0.77

icz

0.75

åĤ

0.73

abl

0.69

çīĪ

0.68

ens

0.67

 worshipped

0.66

ãģ¾

0.65

 profess

0.64

Activations Density 0.040%

pronouns and verbs indicating self-directed actions

references to personal pronouns and the concept of self

No Comments

No Known Activations