INDEX

Explanations

statements indicating various incidents or events occurring in different locations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the verb "to be" in various forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

lycer

-0.77

 pedals

-0.68

inav

-0.66

Premium

-0.63

 contradiction

-0.62

yip

-0.62

 rays

-0.61

 Robots

-0.60

 impulses

-0.60

 Instruction

-0.59

POSITIVE LOGITS

 raided

1.41

 evacuated

1.39

 reopened

1.30

 bombed

1.28

 demolished

1.26

 vandal

1.26

 invaded

1.22

 overrun

1.22

 renovated

1.22

 closed

1.18

Activations Density 0.146%

statements indicating various incidents or events occurring in different locations

occurrences of the verb "to be" in various forms

No Comments

No Known Activations