INDEX

Explanations

references to the name "Sam" with some variations, potentially indicating a focus on specific individuals named Sam

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Sam."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

err

-0.62

 fences

-0.61

 GDDR

-0.60

 troll

-0.59

 Morales

-0.58

 commissioners

-0.57

 trolling

-0.57

_-

-0.56

 Daylight

-0.56

 CONTIN

-0.55

POSITIVE LOGITS

pling

1.32

urai

1.29

ples

1.27

pler

1.18

oan

1.14

uel

1.08

pled

1.03

mons

1.00

plings

0.93

mys

0.93

Activations Density 0.014%

references to the name "Sam" with some variations, potentially indicating a focus on specific individuals named Sam

occurrences of the name "Sam."

No Comments

No Known Activations