INDEX

Explanations

references to the word "rum" at various strengths

oai_token-act-pair · gpt-3.5-turbo

mentions of "rum."

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

lihood

-0.85

 Parenthood

-0.84

KI

-0.65

 Shades

-0.64

Kut

-0.61

 Wildcats

-0.61

 Egyptians

-0.61

 Stafford

-0.61

 Cancel

-0.61

Merit

-0.60

POSITIVE LOGITS

inating

1.01

rum

1.00

oured

0.99

ble

0.98

atis

0.93

pled

0.91

rum

0.85

inally

0.84

inate

0.83

ination

0.83

Activations Density 0.009%

references to the word "rum" at various strengths

mentions of "rum."

No Comments

No Known Activations