INDEX

Explanations

phrases related to the name "Doc"

oai_token-act-pair · gpt-3.5-turbo

references to a character named "Doc"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 proportional

-0.68

uliffe

-0.67

 geographically

-0.66

Rated

-0.65

 suspending

-0.63

mare

-0.62

adding

-0.62

stretched

-0.61

cffffcc

-0.61

 Borough

-0.61

POSITIVE LOGITS

Doc

1.36

doc

1.12

doc

1.08

Doc

1.05

herty

1.03

uments

1.02

iencies

0.88

ually

0.83

sis

0.83

docs

0.82

Activations Density 0.009%

phrases related to the name "Doc"

references to a character named "Doc"

No Comments

No Known Activations