INDEX

Explanations

references to the word "tape", particularly duct tape

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

«<

-0.67

dci

-0.66

 fortn

-0.64

SBC

-0.62

 squa

-0.60

ftw

-0.59

»>

-0.59

 tiffany

-0.58

oun

-0.58

 softshell

-0.58

POSITIVE LOGITS

 tape

1.64

tape

1.51

 tapes

1.40

 Tape

1.38

Tape

1.36

 TAPE

1.16

 taped

1.15

tapes

1.12

 taping

1.04

 Tapes

0.94

Activations Density 0.086%

references to the word "tape", particularly duct tape

No Comments

No Known Activations