INDEX

Explanations

numbers at the beginning of sentences

oai_token-act-pair · gpt-3.5-turbo

occurrences of the number "1."

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

awaru

-0.92

 Lumpur

-0.82

ategory

-0.77

apon

-0.75

usalem

-0.70

thinkable

-0.67

ato

-0.67

unicip

-0.65

orative

-0.65

rador

-0.64

POSITIVE LOGITS

st

1.28

Password

1.19

1.11

1.10

1.05

0.97

0.87

âģ

0.83

0.82

 sqor

0.77

Activations Density 0.093%

numbers at the beginning of sentences

occurrences of the number "1."

No Comments

No Known Activations