INDEX

Explanations

instances in which the word "it" is emphasized

oai_token-act-pair · gpt-3.5-turbo

the word "it" in various contexts

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

-0.77

 Guant

-0.73

Priv

-0.71

 Breast

-0.68

 Transcript

-0.68

 Franc

-0.67

 Friend

-0.67

inqu

-0.67

 Institution

-0.66

priv

-0.66

POSITIVE LOGITS

alian

1.15

chy

1.07

asca

1.04

unes

1.00

 pains

0.90

self

0.88

ain

0.88

 didnt

0.87

 seems

0.87

sterdam

0.86

Activations Density 0.399%

instances in which the word "it" is emphasized

the word "it" in various contexts

No Comments

No Known Activations