INDEX

Explanations

personal pronouns and reflexive pronouns

oai_token-act-pair · gpt-3.5-turbo

references to the concept of self or personal identity

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

 Amend

-0.67

 Sierra

-0.62

 airspace

-0.61

 Clover

-0.61

 racket

-0.59

Vil

-0.59

akings

-0.59

 Noon

-0.58

 Scrib

-0.58

allas

-0.58

POSITIVE LOGITS

selves

1.20

destruct

0.92

ortium

0.91

pecially

0.88

ridges

0.87

self

0.86

theless

0.78

terday

0.77

 explanatory

0.76

contained

0.76

Activations Density 0.019%

personal pronouns and reflexive pronouns

references to the concept of self or personal identity

No Comments

No Known Activations