INDEX

Explanations

words related to the idea of something being deeply rooted or fixed within something else

oai_token-act-pair · gpt-3.5-turbo

concepts related to social constructs and inherent conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

sych

-0.65

avanaugh

-0.62

é¾į

-0.61

jab

-0.61

urai

-0.58

airs

-0.57

otos

-0.56

aird

-0.56

recomm

-0.56

 Gemini

-0.56

POSITIVE LOGITS

 therein

1.03

 deep

0.94

 beneath

0.91

 within

0.90

 deeper

0.86

 firmly

0.82

 inside

0.82

deep

0.78

inct

0.76

 dormant

0.75

Activations Density 0.130%

words related to the idea of something being deeply rooted or fixed within something else

concepts related to social constructs and inherent conditions

No Comments

No Known Activations